The prevalence and correlates of intimate partner violence in Jamaica/La prevalencia y las correlaciones de la violencia entre parejas intimas en Jamaica/La prevalence et les correlats de la violence conjugale en Jamai que.
This study seeks to identify the factors associated with recent physical and sexual forms of intimate partner violence (IPV) using a nationally representative sample of 9,641 women involved in at least one sexual partnership in their lifetime. Women were asked questions about the specific act of physical and sexual abuse by a previous or existing male partner. Five broad categories involving the socio-demographic background, household characteristics, socialization towards violence, agreement with gender norms, and the controlling nature of the male partner were deemed critical in distinguishing women who experienced violent acts from those who did not.
Results from logistic regression analysis indicate that the young age of a woman, belonging to one of the poorest households, and having a controlling partner consistently increased the likelihood of exposure to all forms of IPV, with control being the strongest. We recommend that future research explore dimensions of male control in intimate relationships.
Este estudio busca identificar los factores asociados con los casos recientes de violencia fisica y sexuales entre parejas intimas utilizando una muestra nacional representativa de 9.641 mujeres quienes han tenido por lo menos una relacion sexual en su vida. Se les preguntaron sobre el acto especifico de abuso fisico y sexual por la pareja masculina anterior o actual. Hubo cinco categorias generales incluidas los antecedentes socio-demograficos, las caracteristicas domesticas, la socializacion frente la violencia, aceptacion de las normas de genero y la naturaleza controladora de la pareja masculina fueron considerados importantes para distinguir entre las mujeres victimas de actos violentos y aquellas que no fueron.
Los resultados de un analisis de regresion logistica indican que la edad temprana de una mujer proveniente de un hogar entre los mas pobres y con pareja controladora, eran factores que contribuian de modo consistente a la mayor probabilidad de su exposicion a todas formas de violencia entre parejas intimas, con el factor control siendo el dominante. Recomendamos que en el futuro que la investigacion explore las dimensiones del control masculino en las relaciones intimas.
Palabras claves: Abuso fisico, abuso sexual, Jamaica, violencia entre parejas intimas violencia interpersonal, abuso contra conyuges, abuso domestico.
Cette etude vise a identifier les facteurs associes aux nouvelles formes physiques et sexuelles de violence entre partenaires intimes (IPV) a l'aide d'un echantillon national representatif compris de 9, 641 femmes qui ont eu au moins un partenariat sexuel dans leur vie. Ces femmes ont repondu aux questions sur l'acte specifique d'abus physique et sexuel qu'elles aient souffert de la part d'un ancien ou actuel partenaire masculin. Les cinq grandes categories qui ont ete juges critiques pour distinguer les femmes victimes d'actes de violence de ceux qui ne le sont pas comprennent le contexte sociodemographique, les caracteristiques des menages, la socialisation envers la violence, la conformite aux normes relatives aux sexes et la nature controlante du partenaire masculin.
Les resultats obtenus d'une analyse de regression logistique indiquent que le jeune age d'une femme qui appartient a un des menages les plus pauvres, et le fait d'avoir un partenaire de personnalite controlante augmentaient constamment la probabilite d'exposition a toutes les formes de IPV, le controle etant la plus forte. Nous recommandons que les recherches futures explorent les dimensions du controle masculin dans les relations intimes.
Mots-cles: Abus physique, abus sexuel, Jamaique, violence entre partenaires intimes, violence interpersonnelle, violence conjugale, violence familiale.
In the Caribbean, there has been a distinct increasing trend in violent crime, unlike the decrease marked globally, and the region is ranked the second most violent after Central America (UNDP 2012). This trend is unmistakable in Jamaica, and Harriott (1996) has placed Jamaica currently in the third stage of the development of criminality, typified by an increase in organized crime and gang conflicts along with increases in homicide, offences with firearms and violence against women. The threat of injury, disability, diminished quality of life and health, or death is particularly alarming, as women feel the physical effects and report them to a greater extent than men (Hearn 2012; UNDP 2012; WHO 2010; Barrow 2009; Le Franc, Samms-Vaughan, Hambleton, Fox & Brown 2008). More importantly, the finding that the majority of the reports of violence and injury occur in intimate relationships rather than among other persons known to each other, or among strangers (Le Franc et al. 2008; Krug, Mercy, Dahlberg, & Zwi 2002), has led to a focus on relationship dynamics as the cause of violence between sexual partners in Jamaica and the wider Caribbean. For these reasons, our aim in this paper is to explore the correlates of intimate partner violence (IPV) from available Jamaican data on women currently or previously partnered, and distinguish those factors which place women at increased risk of physical forms of IPV as against sexual forms of IPV.
Since the 1980s IPV has become a serious public health challenge given its impact on women's physical, sexual, reproductive, psychological, financial and social well-being and recognition of the magnitude of the problem (National Family Planning Board 2010; Garcia-Moreno, Jansen, Ellsberg, Heise, & Watts 2006; Hindin & Adair 2002; Sorenson, Upchurch, & Haikang 1996). It has also been identified as a human rights issue as IPV is an infringement of basic rights which should be protected and upheld by nation states (WHO 2010; Garcia-Moreno et al. 2006).
... violence against women [is recognized] as a violation of many human rights: the rights to life, liberty, autonomy and security of the person; the rights to equality and nondiscrimination; the rights to be free from torture and cruel, inhuman and degrading treatment or punishment; the right to privacy; and the right to the highest attainable standard of health (WHO 2010: 6).
The sociological meaning attached to viewing relationships between sexual partners as latently violent has been overlooked in the classical literature on violence (Hearn 2012). The inherently contradictory nature of violence within an intimate setting has been referred to as the "fundamental paradox" (Hearn 2012) and "distant intimacy" (Michalski 2004) in explaining the range of behaviours and definitions that constitute power plays in intimate relationships. The role of agency, and the identification of the "place of violence" in the domination of women by men provides the evidence of social inequality which varies by socio-cultural context. This framing helps us to understand the challenge that IPV poses to population development.
Although studies on IPV from North America, mainly the United States, and Europe dominate the academic landscape (Alhabib, Nur, & Jones 2010; Jewkes 2006), from as early as 1988, the prevalence of IPV in the Caribbean has been documented. Parsad (1988) reported that 83% of the East Indian wives in his study (n=120) had experienced IPV, with 12.5% experiencing it daily. A decade later Haniff (1995) reported that the majority of the 409 acts of violence against women in Jamaica in 1994 were committed by men against their intimate partner. This occurs in an environment where a woman's family, community and the wider legal system discourages reporting (Jones 2003). From studies by Gopaul & Cain (1996) and Peake & Trotz (1999) we gather that there is little variation in the rates of IPV among women of Indian and African descent, and among women in various social classes in Trinidad and Tobago (Kempadoo 2009). Le Franc & Rock (2001) present evidence that in Barbados, sexual violence often follows physical abuse.
From 2005, many of the studies on IPV in the region were population-based following the agendas set in the mid-1990s from the Fourth World Conference on Women (1995) on the heels of the World Conference on Human Rights (1993). These conferences indicated the value of qualitative studies in presenting the "particularities" of women's conditions (Jones 2003) and the need for empirical data to demonstrate the extent of the social problems they faced. This highlighted the need for the development of policies and interventions that are guided by evidence, found to be crucial to emerging public health agendas. Studies by Gage (2005), Gage & Hutchinson (2006) and Le Franc et al. (2008) were some of the Caribbean responses to this call.
The work by Gage (2005) has revealed that, in the year prior to the 2000 Demographic and Health Survey, forced sex was the most common act of violence experienced by 15% of Haitian women in their sample, who had been married at some time in their lives. The prevalence of sexual and physical violence in the previous year (17% in both cases) exceeded that of verbal violence (13%), and 12% of women experienced multiple forms of violence. The levels of violence between intimate partners in the English-speaking Caribbean exceed that experienced by Haiti. The 2003 study on interpersonal violence in the population aged 15 to 30 years of the three largest English-speaking territories (Jamaica, Barbados and Trinidad and Tobago), highlighted the patterns of abuse between partners, distinct from abuse among persons known to each other and strangers (Le Franc et al. 2008). These researchers found that 45% of the female respondents in Jamaica reported physical violence in their relationships, which was similar to the level among women in Trinidad and Tobago (45%) but less than Barbados (50%). However the highest levels of sexual coercion were reported among Jamaican women (73%), with levels of prevalence found to be similar in Barbados (53%) and Trinidad and Tobago (52%).
With the release of data from the 2008 Jamaica reproductive health survey, the National Family Planning Board's report (2009) has provided further evidence of the levels and forms of IPV among women over the course of their lives. This report indicates that 35% of Jamaican women between age 15 and 49 experienced some form of violence from their partner over their lifetime. In fact, 17% reported physical abuse, while 8% reported sexual violence. However, 31% reported verbal abuse. Though it is expected that the prevalence of abuse in the previous year is lower, the recency of the reference point reduces the levels of recall lapse and therefore the likelihood of underreporting. In the year before the survey, the levels of physical (7%) and sexual abuse (3%) reported were concentrated in urban areas, and were exceeded by the levels of verbal abuse reported (15%). The bivariate analysis also revealed that almost twice as many women in common-law and visiting (1) (consensual) unions (20.1%) reported physical violence, compared with married women (11.0%). Higher prevalence of IPV was found among women of with large family sizes, low levels of education, low wealth quintiles, and those who had witnessed violence between their parents (National Family Planning Board 2010: 378379). High levels of disagreement with statements like "a good wife obeys her husband even if she disagrees" (68.6%) and "it is alright for a wife to refuse sex with her husband if he sees other women" (65.4%) seen against low levels of agreement with " it is important for a man to show his wife/partner who is boss" (29.5%) reflect women's unaccepting attitudes to certain displays of male dominance, given their widespread conviction that no circumstance justifies a man hitting a woman.
This information on the prevalence of IPV from the reports of the national survey (NFPB 2010) is useful, but does not supply us with information on the relative importance of various correlates of partner violence which are known to place women at risk in other contexts. Advanced methods are available for distinguishing those factors which place women in jeopardy of experiencing IPV, from those factors which protect women, comparing differences in risk for various population subgroups, (2) and evaluating the likelihood of exposure to IPV for various forms of violence; and it is to these questions that we turn in this paper. We seek to investigate the frequency and risk factors for IPV in the year prior to the 2008 reproductive health survey (NFPB 2010), for women who have had at least one intimate partner. This is critical since the national levels of poverty, unemployment and levels of income inequality remain persistent causes of violent crime and are therefore implicated in interpersonal violence.
In this study five socio-demographic factors identified from previous studies are central to our examination. In exploring women's vulnerability to violence in intimate relationships we focus on the socio-demographic characteristics of the respondent, characteristics of the household to which she belonged, her socialization towards violence, adherence to gender norms and the level of control over her freedom of movement and relationships exerted by her male partner. (3) IPV is depicted by the physical, and sexual violence directed at women in their current or previous sexual relationship. Although the effects of verbal or emotional violence are reported to be more far-reaching than physical violence (National Family Planning Board 2010; Le Franc et al. 2008), and usually accompanies physical and sexual acts of violence, the questions that capture this type of violence are devoid of meaning and motive that would enable us to distinguish between trivial insults and serious behavior to denigrate a partner. Therefore we have not included verbal forms of partner violence in this study. The quantitative analysis is accompanied by a review of selected Caribbean literature.
Recurring themes in studies in various social contexts over the last two decades demonstrate that poverty, patriarchy and male identity, gender norms and the status of women, relationship dynamics and poor conflict resolution skills, alcohol consumption and acceptance of violence are at the heart of the forms and extent and severity of IPV we observe today (WHO 2010; Gage 2005; Gage & Hutchinson 2006; Hindin & Adair 2002; Haniff 1998). Essentially, IPV is the result of a multiplicity of factors at work, and appropriately, ecological frameworks have been used to reflect the hierarchical influence of various risk factors (WHO 2010; Rosenberg et al. 2006; Dahlberg & Krug 2002; Heise 1998; Carlson 1984). In this study, Heise's integrative framework, based initially on Belsky's earlier work (1980) on child abuse, has been adopted to provide the theoretical underpinning to shape our understanding of IPV. Heise identifies the individual, relationship, community and societal level factors that reveal the interrelatedness of the contexts where IPV occurs, while acknowledging the different levels of root causes. Individual level factors include age, income, education, history of substance abuse and previous experience of partner violence. Relationship factors such as poor parenting practices, family dysfunction, marital conflict regarding roles or association with friends who engage in violent behaviors highlight the nature of the family setting for whether women become targets or why men become violent in relationships. The extension of social relationships with neighbours, school, workplace or church reflects the community context which may encourage IPV, while the wider societal factors such as the economic climate, social, health or educational policies that reinforce social inequality or which seek to support violent interactions, are also taken into account. The ecological approach accommodates the insight offered by various disciplines, allowing for explanations beyond male privilege and gender inequality. It therefore helps us to understand the pathways to such violence and the difficulties faced by program planners and policy makers in crafting effective interventions.
Background--The meaning of violence between partners
We are aware that several demographic background characteristics are associated with the risk of IPV. Younger age is associated with higher risk of IPV, while relative age differences of partners have weak effects (Abramsky et al. 2011) and age at marriage has not been shown to have an effect (Jewkes, Levin & Penn-Kekana 2002). Women involved in cohabiting unions are at increased risk of IPV (Abramsky et al. 2011; Brownridge 2004; Yllo and Straus 1981), yet where non-marital unions are uncommon, IPV is associated with marriage (Jewkes, Levin & Penn-Kekana 2002; Ellsberg et al. 1999). Large family size (i.e. high parity) is associated with IPV (Strauss, Gelles & Steinmetz 1980). Although higher education is generally associated with lower risk of IPV (Jewkes 2002), in certain contexts such as the US and South Africa for example, female education has been found in some studies to be protective at both low and high levels of education (Levinson 1989). Poor self-rated health status has also been identified as an important risk factor (Romans et al. 2007; Brokaw et al. 2002). Women are at increased risk of IPV if they work but their partners do not, where both partners do not work, and where they are of lower socioeconomic status (Abramsky et al. 2011; Ellsberg et al. 1999; Gelles and Straus 1998; Bachman and Saltzman 1995). Poverty levels manifested in inadequate household resources and related conflict place women at increased risk. The impact of exposure to violence between parents in childhood has also been shown to increase the risk of IPV in adulthood (Jewkes, Levin & Penkekana 2002; Jewkes 2002; Ellsberg et al. 1999). Furthermore, issues of power and male identity have been implicated in the controlling behaviours by male partners as this is a signal of the skewed balance of power in the relationship, a risk factor for IPV (Jewkes 2002). This is how the international literature has viewed these issues. Of concern to us, however, is the relative importance of these correlates in the Jamaican context i.e., whether Jamaica conforms to global patterns, or presents any differences.
The normative use of violence
When violence is used to settle disputes within families or between neighbors, or when it is used by communities to gain political attention or used for state oppression of certain groups, IPV is more prevalent. Furthermore, where violence is experienced as a child through parents or other adults in the home, the child learns the acceptability of violence. Males are highly likely to be abusers themselves later in life and girls are likely to be victims of physical abuse by partners. According to Jewkes "men learn to use violence and women learn to tolerate it or at least to tolerate aggressive behavior" (2002: 1426).
That Jamaica is characterized as a place of violent interactions and methods of conflict resolution remains evident from reports by Amnesty International (Amnesty International 2007). Poor responses from legal and criminal justice systems, have in the past contributed to the acceptance of violence. However, the Caribbean Human Development report for 2012 indicates that community acts of intervention should be taken into account as we review the indifference to violence against women expressed in condemning attitudes of some opinion leaders (UNDP 2012: 40). However, as a party to the Convention on the Elimination of all forms of discrimination against women (CEDAW), the Inter-American Convention on the Prevention, Punishment and Eradication of Violence against Women, and the Rome Statute of the International Criminal Court, the government has made a clearly demonstrated commitment to implementation of the related legislation and incorporation of international standards. This includes the revision of procedures for crimes of violence against women, the treatment of victims and witnesses, reparations for victims, gender balance of staff and expertise of staff (UNFPA 2009; Amnesty International 2007).
Women as compliant participants
The perceptions that women hold about wife beating also contribute to violence in relationships (Parsad 1988).
Although women in the poorest of nations are probably most inclined to believe that men are justified in beating their wives, in all settings, in developed and developing countries, abused women tend to hold more beliefs which justify violence against them. (Fagan & Browne 1994).
The majority of the East Indian women in Parsad's study (52.5%) considered wife beating acceptable and blamed themselves for the acts against them. This Parsad attributes to their socialization pointing to the fact that women of African origin were more likely to defend themselves (1988: 53). Reports of female perpetration of violence are not absent from the Caribbean literature as Le Franc et al. (2008), Le Franc & Rock (2001) and Haniff (1995) present this evidence. When it becomes customary for partners to treat each other as opponents or rivals, women may retaliate, or report sexual coercion (Le Franc et al. 2008). They may also act in self-defense. On the other hand, other studies have shown that IPV is tolerated by women because of their experience of serial partner abuse, or the intergenerational transfer of attitudes and behaviors from parents or older siblings, who themselves were beaten as children or beaten in their relationships, numbing their own responses in adulthood (Le Franc & Rock 2001). The tolerance is also reported as an expression of the belief systems of each partner. When women believe that a partner communicates his love for her or his need for behavior change through IPV, any expression of violence will be accepted (Le Franc & Rock 2001). This applies to the physical abuse of children as well. In scenarios where abuse is inflicted, many women endure it if the economic support provided by men is not withdrawn (Le Franc et al. 1996).
As a result of this review we now look at the factors that are associated with IPV given the patriarchal basis of men's power, their perspective on how this power is operationalized in relationships, and how this is juxtaposed against women's view of male domination; i.e., their economic dependence on men, and their family building goals. This review helped to inform our selection of variables for inclusion in the analysis which follows. We explore how the characteristics of individual women may engender conflict in relationships thereby affecting women's risk for experiencing IPV.
Power and male dominance
Gender norms and the views that are held by men and women about their roles in society and in their relationship sharply influence whether violence exists in intimate relationships (Doe 2000), especially if discordant views are held by partners in intimate relationships. When men hold traditional views about their roles as breadwinner, and exert power over household members and in the decision-making related to household expenditure, violence is commonplace. When spouses or partners do not share these views or do not acquiesce to male domination or do not accept acts of control which place limits on their sexuality, freedom of movement, household role, resources and morals, abuse between partners is a likely outcome (Jewkes 2002; Chevannes, 2001; Le Franc & Rock 2001).
It is not difficult to agree that to ... remain a man requires certain kinds of interactive relationships with women, ...--sexual intercourse; sexual initiative; special understandings about who does what within the household; who is free to go where and when; moral and material control, and so on (Chevannes 2001: 218-219).
Thus relationship conflict brought about by the "transgression" of gender roles or "challenges" to male dominance will increase the likelihood of violence in relationships. Further themes derived from the Caribbean literature elucidate the ways the threat to male dominance is played out. Men's fear of a partner's infidelity, real or perceived, arises out of the disrespect to his power, authority and threat to ego (virility) that this behavior engenders (Anderson 2012; Gibbison 2007). In Anderson's (2012) recent study on masculinity, a half (51%) of the sample of 1,142 fathers from four communities in Jamaica indicated there had been times when they hit their partner. This was especially so in the inner-city community (73%), where residents were primarily self-employed and the commitment to macho values was the highest (55%) of the four communities studied. Even in the middle-income community, the percentage reached 39%.
However this is not simply the case of a man wanting to control the sexual behaviour of his partner ("ensuring monogamy and conformity") in response to the threat to "traditional arrangements of gender and sexuality". Jealousy is the first of the proximal determinants of IPV identified by Le Franc & Rock (2001) in their presentation of the statements from women they interviewed, and has been documented in several studies (Dobash & Dobash 1997: 268). Goetz et al. (2008) have established that male jealousy due to infidelity of partners is mainly due to the sharing of sexual intimacy (rather than emotional) which occurs and is substantiated through mate retention studies. The attractiveness of the female partner in a relationship and the premium accorded the securing of this male prize is only one part of the story. The reproductive potential of the female partner is far more telling. According to these researchers, the "sensitivity and distress" experienced as a result of infidelity are linked to the "cost" men pay in providing for children who are not biologically theirs, and who may be "genetically related" to a male rival, and the threat to his reputation if this is exposed (4) (Gage & Hutchinson 2006; Platek & Shackelford 2006). This question of paternity of children can cause violent reactions in men. Community discussions held prior to the 1993 study of relationships between men and their families by Brown, Anderson & Chevannes, indicated that:
Most men fear being given a "jacket" (being named as father when the actual paternity rests elsewhere or cannot be determined). While some men may accept a jacket knowingly if it remains a 'family secret' others said they would feel suicidal, or would beat up the woman and break up the relationship (1993: 62). (5)
Therefore IPV is used by men to punish wrongdoing (Dobash & Dobash 1997), and to restrict the women's sexual behavior so as to remove any possibility that the offspring she has is not genetically his. However, this relationship is mediated by the presence of male relatives or economic dependency of the woman (Goetz et al. 2008). This "male proprietariness" Wilson & Daly (1992) argue is related to male entitlement, and broadened to include ownership and control of resources; this extends to ownership and control over a partner's body and sexuality. Given Jamaican women's inclination to having many different partners (and children) to supplement their individual and household economic means and reduce their dependence, men's suspicion of a woman's behavior is understandable (Le Franc et al. 1996). However men may also have multiple partnerships (Chevannes & Mitchell-Kernan 1992), and arguments about the diversion of resources to other women or children so that "pooled" resources are inadequate, or the woman's attempt to negotiate condom use in an environment of perceived unfaithfulness, can prompt violent responses from either partner. The presence and treatment of children in the household, particularly those of a previous partner, is also a source of tension that escalates to violent acts in some relationships (Le Franc & Rock 2001). Conflicts over inadequate household resources and control over their distribution in Jamaica, (Le Franc & Rock 2001; Le Franc et al. 1996), the payment of dowry in South India (Rao 1997) or alcohol use in South Africa (Jewkes, Levin & Penn-Kekana 2002) are also powerful determinants of IPV.
It is clear that although women may play the dominant role in the domestic sphere (Chevannes 2001) women who are dependent on men for the economic viability of the household are most vulnerable to violence (Le Franc et al. 1996). Thus working women, women who are financially independent, or who are better educated are at lower risk of experiencing IPV as they are actively able to contribute to their own economic well being and that of their children, and remove themselves from situations where this wellbeing is threatened.
However, these same factors considered to be protective, may elevate the risk of exposure to IPV when the improvement in women's status (income, education, occupation and social roles) results in having more power over the resources available to the household or wanting to have more say in decision-making, or even wanting to leave the relationship (DeKeseredy & Schwartz 2005), as this "may promote male insecurity and feelings of economic inadequacy, leading to more violence in relationships" (Hindin & Adair 2002; Jewkes & Abrahams 2002). This explains why women's economic vulnerability is linked to violence. A statement from focus group discussions held in the mid-1990s with female Informal Commercial Importers and Free Trade Zone workers in Kingston reflected this sentiment:
When a woman inside, and is them [i.e. the man] bring in the bread, them feel like them supposed to wipe them foot on you. The man think when you're working them don't have that hold on you (Le Franc et al. 1996: 1415).
The transition in women's economic power relative to men's is evident where the highest levels of IPV are reported (WHO 2010). It is therefore the shift in the balance of power in relationship, regardless of the cause, brought about by women's economic empowerment in a situation where a man's own status fails to live up to his notions of masculinity and successful manhood, that promotes the environment for partner violence, as men attempt to recover the power balance (Jewkes 2002; Gelles 1974). When men cannot support their partners adequately or at all, through unemployment and poverty, they cannot manage their domestic arrangement and as a consequence cannot control their partners. This Le Franc and Rock (2001) describe as "patriarchy without power".
These themes uncovering the association between male dominance and IPV are repeated in the literature across the world. Very often, efforts to empower or improve the status of women have ambiguous effects, in some cases being protective and in others, increasing IPV (Bates & Schuler 2004; Jewkes, Levin & Penn-Kekana 2002; Schuler et al. 1998; Schuler 1996).
The study has two main sections. Firstly, univariate and bivariate analysis is undertaken to provide a description of the main characteristics of the sample. This precedes the multivariate analysis in which three models are proposed to predict the likelihood of experiencing all forms of partner violence, physical and sexual partner violence. Of concern to us, and therefore informing our hypotheses is whether higher rates of IPV of any type, or partner violence of a physical or sexual nature, are to be found among women of young age, involved in a common-law union, having a small family size, lower levels of educational attainment and wealth status, poor health status, working, having childhood exposure to violence in childhood, predisposing attitudes to partner violence, and involved with a controlling partner.
The data utilized are from the 2008 reproductive health survey (NFPB 2010) administered through face to face interviews to a nationally representative sample of women from various conjugal union types between age 15 and 49 years. The survey covered topics on women's pregnancy, birth and reproductive histories, general health status, partner history and family life, sexual activity and contraceptive use, knowledge of HIV transmission and risk status, along with other socio-demographic background factors. The questions on gender violence and norms were added for the first time in 2008 because of the widespread observation of factors linked to partner violence, (6) increasing reports of this behaviour, and the link made between violent behaviours among intimate partners and negative reproductive health outcomes (7) (National Family Planning Board, 2010). Women who indicated that they had experienced abuse in the year prior to the survey were asked questions about the specific act of physical, sexual and psychological abuse by a previous or existing male partner i.e., their husband, ex-husband, any other man they were living with or had a visiting relationship with. Questions modified from the revised Conflict Tactics scale (CTS2) (Straus et al. 1996) were designed to facilitate identification of the population sub-groups most at risk, and draw attention to the factors which placed them at risk, so that ultimately the number of new cases is reduced through programmes and policies informed by scientific inquiry. The CTS2 obtains data on the physical and psychological acts of abuse inflicted by one partner in a casual, cohabiting or marital relationship, against the other partner, and has been tested and validated in Jamaica (Le Franc et al. 2008). In addition, the controlling behaviours of their partners, their attitudes towards male domination, the experience of abuse in childhood and as an adult outside of the partnership were observed.
The sample for this study consisted of 9,641 women of reproductive age who had been involved in at least one sexual partnership in their lifetime. The data reflect cases weighted so as to estimate error/changes to sample selection and non-response. The statistical package for the social sciences (SPSS) software was used for analysis.
The Jamaica Reproductive Health Survey 2008 (NFPB 2010) is a population-based probability survey consisting of face-to-face interviews with women (15-49 years). (8) Using the 2001 census as the sampling frame, the first stage of the three-stage sample design was the selection of Enumeration Districts (EDs) and all 14 parishes of Jamaica were further subdivided into 307 "sampling regions" of approximately equal size, which constitute the strata for the RHS sample. Within each sampling region EDs were selected with probability proportional to the size of the ED, which is measured by the number of households in the ED, according to 2001 census definition. A total of 628 EDs were selected as primary sampling units (PSUs) and an updated count of households in each PSU was obtained. The second stage of sampling involved the random selection of clusters of households within each PSU chosen in the first stage. Separate households were selected for male and female interviews. Determination of the cluster sizes was based on the number of households required to obtain the target average number of completed interviews per cluster in each parish. The total number of households selected in each cluster took into account information from the previous survey (2002) on the proportions of unoccupied households, households with women aged 15-49 years. In the third stage, one woman between the ages of 15 and 49 was selected at random for interview from each of the households selected in the female sample.
The main outcome variable used for logistic regression analysis was abuse experienced in the year prior to the survey. The indicator of physical violence was taken from six questions on whether women had been slapped or had something thrown at them; whether they were pushed, shoved, or had their hair pulled; whether their male partner had hit them with his fist or something else that could hurt her; whether her male partner threatened her with a gun, knife or other weapon; had choked her or burnt her on purpose; and had kicked her, dragged her or beat her up. Sexual violence exposure was measured by one question on whether the woman had been physically forced to have sexual relations when she did not want to. Current prevalence is the proportion of women who had been engaged in at least one partnership reporting at least one act of psychological, physical or sexual violence during the twelve months before the interview. Therefore, the physical abuse variable was coded "1" if a woman reported she was slapped or had been choked by her partner within the past 12 months. This variable was coded "0" if the respondent indicated that she was neither slapped or choked within this time period. An additional variable was created to reflect multiple forms of violence, with no violent acts experienced coded "0", the experience of one act was coded "1", and a code of "2" when two or more acts were reported.
The study also looks at controlling behaviours by male partners. Six behaviours were reported on, and this includes acts to restrict a woman's contact with family and friends, limit her mobility, treat her indifferently, the partner's anger when she speaks to another man, his suspicion of her unfaithfulness, and that she seek his permission for seeking healthcare. Women who agreed with one or two statements about the controlling nature of the partner were categorized as experiencing low levels of control in their relationship, while those agreeing with three or more statements about their partner were treated as experiencing high levels of control in their relationship. Respondent attitudes to gender norms were also considered important for inclusion. This was measured through agreement with six statements of currently held beliefs about how men should treat women and how women should respond to men, within families. In tandem, situations justifying women's acceptance of physical violence were explored. Failure to complete household work, disobeying the male partner, denying him sex, confronting suspicions of his infidelity and her infidelity, were the situations investigated.
Witnessing violence between parents or step parents, or experiencing punishment from a parent or family member, or the experience of violence before age fifteen has been reported to inculcate the norming of violence, increasing its acceptability so that persons with this history are more likely to inflict violence themselves and be victims of IPV (Le Franc & Rock, 2001). If the respondent responded positively to any of the questions on their history of witnessing or experiencing abuse, a dichotomous variable was created and coded "1". In all cases the reference category was coded "0".
Aside from controlling acts of partner, gender norms, acceptable terms for justification of violence and the witness and experience of violence before age fifteen and experience after age 15, other variables were included in the analysis as co-variates. The individual factors chosen were current age (15-24, 25-34, 35 years and older), parity (childless, 1-2 children, 3 or more), union status at the time of the survey (legal marriage, common-law, visiting, no steady partner), education (9) (8years or less, 9-13 years, 14 or more years) employment status (10) (working or not working), self rating of health status (poor, fair, good or better), and the number of partners. The variables related to the status of the household were urban residence, and wealth quintile, which is a measure of the living standard of the household compared with others. (11)
Multiple logistic regression was used to develop models to examine the associations between the selected correlates and the experience of each form of IPV in the year before the survey. A model was developed for each type of violence in order to identify significant correlates of physical violence, distinct from sexual violence. The construction of scales for the controlling behaviours of the partner, women's justification of IPV and acceptance of gender norms, was not feasible in the effort to focus on the extent of respondent agreement with the individual items which comprised these 'latent constructs'.
For the female sample 8,259 women from 8,542 households included at least one eligible respondent aged 15-49 years resulting in a response rate of 96.7%. As many as four visits were made to each household with eligible respondents who were not at home during the initial household approach. Almost all respondents who were selected to participate, and who could be reached, agreed to be interviewed. Less than one percent of eligible women refused to be interviewed, and 2.5% of women could not be located. Even though the overall response rate was similar in urban and rural areas, eligible respondents in urban areas were somewhat more likely to refuse to be interviewed.
Firstly, univariate analysis was undertaken to provide a description of the main characteristics of the sample. Bivariate analysis was then used to estimate crude associations between each measure of exposure and IPV for women exposed in the year prior to the survey to physical and sexual violence. This precedes the multivariate analysis in which three models were proposed to predict the likelihood of experiencing any form of partner violence, physical and sexual partner violence. The focus was to unearth the similarities and differences in the patterns of associations for each form of violence. In reporting results, we consider odds-ratio between 1.05 and higher as being risk factors, while odds ratios of .95 and lower were protective. Statistical significance is considered at the 5% level.
Sample Characteristics--Univariate and Bivariate results
The socio-demographic characteristics of the sample are presented in Table 1. While about a quarter of all women were in the youngest age category, over a third were in the 25-34 age group (36%), with 39 percent in the oldest group. Of the four union types, the majority of women were either in visiting unions (37%) or common-law unions (27%), while women in more stable marital unions reflect the lowest proportion, (17.6%). About one-fifth of the sample was with no steady partner. Although 21 percent of the women in the sample had no children, over twice this proportion (46%) had one or two children, and 36% had three children or more. The majority of women (64%) were reported to have achieved a secondary level of education and most were found to be living in urban areas (58%). In the survey year, about half the sample were working (49%), while those not working (51%) included 25% who were keeping house, and 14% who were unemployed. Students represented one in every 20 women, while one in 15 women accounted for those who were home and not keeping house. Approximately 41 percent of the women resided in households falling in the two lowest wealth quintiles i.e. the poorest households, while 59 percent of women lived in households considered to be middle income or better. Of the sample, about 83 percent had experienced at least one act of IPV in the previous year.
In reviewing the characteristics which distinguish women who have experienced IPV from those who have not, we unearthed those factors which increase women's vulnerability. In 2008, 28 percent of the women experiencing IPV were found in the 15-24 age group, with 42 percent in the 25-34 category, and 31 percent for women age 35 years or older. Of the women who did not experience IPV, 59 percent were found in the two younger age groups. It is only among the oldest women 35-49 years that the proportion of women with no history of IPV (41%) exceeded that of women with the experience (31%). While only 15 percent of those experiencing IPV were in a marital union, more than twice this proportion were found in common-law (37%) and in visiting unions (36%), and 12.2% were without a partner.
We also found that 28 percent of those experiencing IPV had no more than nine years of education i.e. no further than primary schooling, compared with 21 percent who had no experience of IPV. The disadvantage of women experiencing IPV relative to women without the experience extends to their employment status and levels of wealth. About 46 percent of women who experienced IPV were working, while 50 percent of women without this exposure were working. And, over a half (53%) of the women experiencing IPV were found in households representing the two lowest categories of wealth, compared to 39 percent of women who did not experience IPV. It is also clear that the largest proportion of women exposed to partner violence had either one or two children (43%), compared with 26 percent of those not experiencing IPV. Little difference was found in the pattern of residence between women experiencing IPV and those without this experience.
A review of selected health characteristics of women in the sample revealed that there was no major difference between women experiencing IPV in the previous year, and those who did not, in terms of the proportions who were currently pregnant, or who had used contraception in the last year. This is found in Table 2. However, with respect to age at first sex we found that one-fifth (21.1%) of the women who were victims of partner violence experienced their sexual debut before age 15. Only one-eighth (12.1%) of the women who were not victims of partner violence experienced their sexual initiation before the age of 15. Of the women who did not experience partner violence, a much larger proportion (13.7%) had initiated their sexual experience at age 20 or older, compared with 6.1% of the victims of abuse. We also found that a greater proportion of victims of abuse rate their health negatively when compared with non victims. About 38.1% of the women who were victims of violence rated their health as either fair or poor, compared with 25.4% of those without exposure to IPV.
The experience of violence in childhood is one of the most pervasive determinants of IPV and is presented in Table 3. We have found that 30.9% of the women who had experienced IPV had witnessed one parent hit, slap, lick or shove the other, before age 15. This compares with 17.3% of women without exposure to IPV. We also gathered that the vast majority of women with recent experience of IPV (81.0%) had been physically hurt by a parent or other family member prior to their fifteenth birthday, while 68.5% of women with no exposure to violence were faced with this early occurrence. This is found in Table 4. Beyond age 15, it is clear that women experiencing partner violence had been hurt by someone other than their current or past partner. About 23% of the women spared the experience of IPV had been physically hurt by someone other than their current or previous partner, while twice this proportion (45.4%) of the victims of IPV had experienced violence from an acquaintance, family member or stranger in their late teens or after.
The controlling nature of the respondent's partner is also highlighted in the literature as an important determinant of the risk of experiencing partner violence. This is evident from the data in Table 5. As expected, the proportions of women experiencing IPV and agreeing with the statement about the behaviour of her partner exceed the proportions of women without a history of IPV, in every instance. Of the women who were victims of IPV, 45.4% had partners who suspected their unfaithfulness, 57.2% had partners who controlled their mobility, while 60% had partners who became angry if they spoke with another man. For women without a history of IPV the proportions were 14.9%, 30.5% and 24.2%, respectively. The greatest disparity between the two groups of women (i.e. between victims and non-victims of IPV) is found in the response to the statement about how the male partner behaves when she speaks to another man.
Although the evidence presented in Table 5 would suggest that IPV is endemic to the culture, we have found that the proportion of women agreeing to statements justifying violence between partners is minimal. This is presented in Table 6. It is only in the event of the woman's infidelity that 3% of women experiencing IPV and 2.3% of those without this experience believe that violence between partners is justified. Our review of the responses to gender norms related to the roles to be played in intimate relationships, presented in Table 7, indicates there are no large gaps in the proportion of women who uphold positive beliefs about their role, when we compare women experiencing IPV and those without the experience.
While the effect of age is commonly found in the literature, the effect of wealth quintile is not. If we review the acts of violence perpetrated by male partners according to the wealth quintile of the household to which women were attached, we find that the proportion of women reporting acts of violence steadily declines the better off a woman is. This is presented in Table 8. About 22% of the women in the lowest quintile had experienced some form of violence, while about 10% of their counterparts in the highest quintile reported exposure. For women indicating physical abuse, 8.6% of those at the lowest quintile reported being victims of a physical act, while 9.2% of the women in the next to lowest quintile had faced such an incident. Only 3% of the women in quintile five had experienced this. In the case of sexual abuse, women in quintile one and two had similar levels of exposure, with 4.2% and 4.1%, respectively; but only 1.4% of the women in quintile five. On average women in the three lowest quintiles experienced one act of violence, and this was usually of a physical nature. This is presented in Table 9.
Findings from multivariate analysis
Exposure to any form of IPV
We have found that there are statistically significant differences between women who have been exposed to IPV of any kind, and those who have not. A woman's age, negative health rating, and the number of lifetime partners she has are the demographic factors associated with the increased likelihood of IPV in the year prior to the survey. Though women aged 15 to 24 years face odds of experiencing IPV (1.8) that are similar to women aged 25 and 34 years (1.8), those women under age 35 years faced a higher likelihood of IPV than women age 35 years or older. Women providing more unfavourable ratings of their health faced a greater likelihood of experiencing IPV. Women reporting fair health were 1.4 times more likely to experience IPV compared with women who reported good health. Women reporting poor health were slightly more likely (odds ratio=1.5) to have encountered IPV. As expected, the greater the number of lifetime partners, the higher the likelihood of being a victim of IPV. However, we were surprised to find that being involved in a common-law union did not significantly affect the experience of any form of IPV, yet women in visiting unions and with no steady partner were less likely than married women to have this experience.
From the household characteristics we found that residence in an urban household or a poor household raised the chance of encountering IPV. In fact, compared with wealthy women (quintile 5) women from the poorest households (lowest, quintile 1) were 2.3 times more likely to experience IPV, while women from the third quintile were 1.7 times more likely to experience IPV. Women in quintile 4 were not significantly different from the women in quintile five.
The findings related to a woman's socialization towards violence, and the controlling nature of her partner indicate that these factors place women at risk of IPV. Women who witnessed violence between parents during their childhood were 1.3 times more likely to have experienced IPV. In like fashion, women who were hit slapped or kicked by a parent or other adult family member were 1.4 times more likely to report IPV. And women who experienced violence from a non-partner (relative or stranger) after age 15 were 1.7 times more likely to have experienced partner violence, compared with women who did not have this experience. However, it is the impact of the controlling nature of the partner that is most telling. Women who reported low levels of partner control were 2.9 times more likely to have experienced IPV, relative to those without controlling partners. However, women with partners who were highly controlling were far more likely (8.1 times) to have experienced IPV.
The analysis also revealed that the factors reducing the odds of experiencing IPV were adherence to gender norms, and being in one of the more unstable or non-residential unions. Women who expressed agreement with two or three statements reflecting gender norms were 1.4 times less likely to have experienced IPV, while women agreeing with four or more statements were 1.6 times less likely. The model developed to distinguish the characteristics of women who experience IPV explained 27 percent of the variation in IPV of any type, and correctly classified 83 percent of the sample.
Exposure to physical forms of IPV
Our findings further reveal yet a very different set of characteristics of women with partners who have been physically violent. We have found that a young age, being in a common-law union, residing in an urban area, belonging to a household falling in the lowest three wealth quintiles significantly increase the risk of exposure to physical violence. Compared with women 35 years or older, the youngest women (age 15-24 years) were three times more likely to experience a physically violent act initiated by their partner, while women aged 25-34 years were twice more likely. While women in common-law unions were 1.6 times more likely to have had a physically violent encounter with their partner, those in visiting relationships or with no steady partner were not statistically significantly different from married women. Relative to women living in rural areas, those living in urban areas were two times more likely to have been physically abused by an intimate partner. Women in each of the three lowest quintiles were two times more likely than wealthy women to experience physical violence.
Of the socialization factors, the experience of violence after age 15 exerted the strongest impact. If a woman was slapped, kicked or shoved by someone other than a current or former partner, after the age of 15 years, she was three times more likely to be hit by her partner. Again, the partner's control in the relationship played an important role in the experience of physical violence. Women who described their partner's control as weak were 2.5 times more likely to experience physical IPV than those whose partner were not controlling. However, women whose partners extremely controlling were ten times more likely to experience a physically violent act. The single protective factor was adherence to the prevailing norms related to the roles and behaviour of men in relationships.
Exposure to sexual forms of IPV
The model developed to reveal the factors associated with the experience of sexual forms of violence between partners has shown that beside age, having no more than a primary school education, having less than three children and having two or more partners over the course of their lives, placed women at a severe disadvantage. In each instance, women were two times more likely to experience a sexually violent act, compared to women in the respective reference categories. Comparable odds ratios were found for women residing in urban areas, and women residing in the poorest households.
The surprising finding is that none of the factors reflecting the socialization towards violence or adherence to gender norms were found to be statistically significant in relation to sexual violence. Yet the element of control between partners, as in the models looking at physical violence between partners, had the strongest effect of all correlates in this model, and the greatest effect across the three models of IPV. Women with low levels of control exerted by their partner were three times more likely to experience a sexually violent act, but those with a highly controlling partner were 12 times more likely to have this experience. Being in a non-marital union and working lowered the risk of exposure to this form of violence.
Our findings indicate that the characteristics predicting exposure to any form of violence in the last year may differ when we focus on each form of violence between partners; this difference is seen in the consistency and strength of impact of the correlates that were analyzed. Being of a young age i.e. lower than age 35, belonging to a poor household (wealth quintile 1 and 2), and having a partner who displays controlling behaviours, consistently places women at risk of experiencing all forms of IPV perpetrated by a partner, that we have explored.
We have also gathered that a woman's socialization towards violence and acceptance of gender norms are not steady predictors of her exposure to IPV. Those women who have witnessed it in childhood or experienced violence at the hands of a parent or other family member are more likely to experience any form of abuse. But the experience of violent interactions after the age of 15 years from someone who is not a partner, but who may be a relative, friend or stranger, makes her vulnerable to the risk of any form of IPV and physical acts of violence inflicted by a partner. Conformity to gender norms has the effect of increasing the likelihood of experiencing any form and physical forms of IPV, rather than sexual forms.
Of the background variables we have analyzed, we recognize that education has a limited effect on the experience of IPV, for it is only the women with no more than primary schooling who are more likely to experience sexually violent incidents, than those with tertiary-level training. This is not so when we look at the experience of IPV within unions. It is clear that women in common-law unions are more likely than married women to experience physical acts of violence, but along with those in visiting unions, are less likely to be exposed to sexually violent acts, compared with married women. While a woman with less than three children has an increased risk of exposure to sexual violence relative to women with larger family sizes, women who work are protected. The analysis has also revealed that urban residence places women at risk of physical and sexual forms of IPV, but having more than one partner predicts exposure to sexual forms. A woman's assessment of her health status as being either fair or poor increased her chances of any form of IPV.
In this study we have sought to highlight the impact of several correlates known to predispose women to acts of violence within an intimate partnership. The findings are striking for several reasons. They provide, for the first time, the statistical evidence of the impact of several correlates, and the unique set of risk factors for each form of violence. Higher rates of prevalence can be found among poor women living in urban areas, who are young, have little education, and who are involved in common-law relationships, who experience violence outside their intimate relationships, who are living in relationships where the male partner is controlling. These impacts are not surprising, but their value lies in the complete information about Jamaica that has been provided.
The cross-sectional data upon which this study is based provides us with a snapshot of the prevalence of IPV at one point in time, ignoring the temporal nature of the parameters of concern, best investigated in longitudinal designs. We are therefore not able to establish causality. Women's self reports of IPV may reflect under-reporting by women (Haniff 1998), changing the relationships observed and weakening the associations found (Hindin and Adair 2002). Although the perspective of the female victim used in this study is the most frequent approach used, (which assumes that men are the perpetrators), we acknowledge that reports of victims and perpetrators, male and female alike, should be used to balance the analysis. The divergence in the accounts of IPV from men and women warrants this approach (DeShong 2011; Dobash and Dobash 2004; Le Franc and Rock 2001), and is missing here. We are also unable to analyze the role of neighbourhood characteristics such as the percentage living in poverty or unemployed, or in certain occupations, due to the unavailability of these data through population-based surveys.
Le Franc et al. (1996) describe the heightened vulnerability of women who are dependent on men in situations where their combined resources are inadequate, as facing a triple threat. These men under the pressure of their own beliefs of male dominance, and without the means to live up to these beliefs, are prone to take violent action (Le Franc et al. 2008; Chevannes 2001). This study has shown that this action in all its forms is taken against women perceived as weak. The women most at risk are those of young age and who are poor. These are women with rudimentary notions of the way relationships work, seemingly powerless to change the outcomes of previous generations of women; women with underdeveloped ideas of female autonomy, unaware of the realistic ways this can be achieved in a relationship; women living within a culture that exerts pressure to fulfill the vision for their lives through forming relationships and building families based on traditional roles, but without the social structure that gives stability to unions should the roles shift in any way.
Weak women are also those with little education, but we recognize that the effect of education is inconsistent. Although we find no significant effects of education on the risk of exposure to physical forms of IPV, women with no more than primary level education face twice the risk of experiencing sexual violence when compared with those with education beyond the secondary level. This supports the findings from studies showing that low educational attainment is associated with high levels of violence (Jewkes, Levin & Penkekana 2002). However, this finding is incongruent with the analysis of Haitian data in which women with some primary schooling had a higher risk of sexual abuse than uneducated women, while those who completed primary school were no different from uneducated women (Gage & Hutchinson 2006). Such patterns may be related to the inverted U effect where "women having some education empowers women enough to challenge certain aspects of traditional sex roles, but that such empowerment carries an increased risk of violence until a high enough level is reached for protective effects to predominate" (Jewkes 2002).
The extent of income inequality also adds to the complexity of the interaction of factors that place women at risk of IPV, and this is evident in the proportion of poor households with the least resources to satisfy basic material needs, placing greater strain on emotional needs of intimate partners. This helps to explain why the women in the poorest households (lowest and second to lowest wealth quintile) were more likely than women in the wealthiest households to experience both forms of violence. These findings are in tandem with studies showing the prevalence of violence within families being greater where poverty and income levels are low, and where unemployment is high (Gelles 1974; Johnson 2000; Lupri 1990). Kempadoo's work (2009) in Trinidad and Tobago hinted at this through the variation in the rates of IPV among women of differing social classes. The explanations center on the fact that "inequality and relative deprivation are the main causes of violence not because these people are essentially more violent but because of the associated feelings of shame and worthlessness connected with the treatment of people in these circumstances" (Gilligan 2001). Therefore violence in intimate relationships is a part of the reaction to the burden of poverty and limited means of coping, in general (Jewkes 2002). This becomes clear as conflict is minimized when the source of financial support for the poor household lies outside of the male and female partners (Jewkes 2002), though the evidence from Thailand negates such assertions (Hoffman, Demo, & Edwards 1994).
That the strongest factor associated with the three forms of IPV is the controlling nature of the male partner uncovers the challenge men face in intimate relationships when they cannot adequately satisfy the economic needs of the household, or maintain the loyalty of their partner, while these remain markers of their masculinity (Dutton 1996; Ellsberg et al. 2000; Le Franc & Rock 2001). Rather than using more cognitive means related to better methods of communication, or conflict resolution to protect against the threat to their own reproductive potential, cope with anxiety over possible abandonment by a partner or the challenge of economic circumstances, men resort to influencing a partner's behaviour through regulating her movements, limiting her association with family and friends to produce compliance, regain the power they believe they should have, and reestablish dominance (Gage & Hutchinson 2006).
Though Felson & Messner (2000) argue that, outside of their study, no statistical evidence of the impact of control exists, their analysis showed that the occurrence of violence which involves control, through the issue of threats made prior to a violent incident by the male partner, exceeds the occurrence in violent attacks involving antagonists. The Felson and Messner study assumed that physical harm would take place if there was no compliance with the demands of the threat, thereby reflecting both the coercive power of the partner and a desire to exert influence in the relationship.
We have also found that women who have witnessed IPV in childhood or experienced violence at the hands of a parent or other family member are not likely to experience physical or sexual partner violence. This is a surprising finding given that existing literature indicates that direct exposure to violence in childhood can result in trauma in witnessing the event, and vulnerability due to "insecurity in attachment which underlie working models of self and other that begin to develop within the earliest caregiving relationship" (Levendosky, Lannert, & Yalch 2012: 422). This can lead to increased risk of involvement in violent relationships either as victim or perpetrator (Le Franc & Rock 2001). However, we have found that women who have experienced violence after age 15 at the hands of someone other than a current or previous partner are likely to experience physical forms of abuse in an intimate relationship. This is attributed to the poor patterns of relating in previous relationships that are carried over to current relationships, and which can lead to intimate relationships in which they have a poor understanding of self and see themselves as victims. Herein lies the tragedy of IPV. Exposure either in childhood or adulthood leads to damaged ways of relating in intimate and other relationships "that can place multiple generations at increased risk of vic-timization and psychopathology." (Levendosky, Lannert, & Yalch 2012: 423).
Any attempt to design interventions and channel resources to address IPV should take account of the structural factors at work, placing priority on youth, and the poor, and should address relationship skills and choices. We may never understand why women choose certain partners, but we are aware that in the attempts to form relationships and build families in less than buoyant economic conditions, convenience and expediency may override in the selection of mates. Perhaps this is the one area where we could shape interventions immediately. Without the necessary data to tailor interventions to reverse the trends in the prevalence of IPV we will continue to be faced with issues jeopardizing individual self efficacy. However, the recommendation to call for a special national survey on Intimate Partner Violence (IPV) is not practical given the links that must be made to the reproductive health and other related aspects of women's lives that bear on their exposure to IPV.
To this end we advocate that future reproductive health surveys capture the gaps in income and education between partners indicative of differences that would undermine the man's position with his partner, and in the household. To this we would add indicators of household decision-making and female dominance, other dimensions of control and coercive power in relationship, and more fully explore personality profiles, external factors that lead up to the violent incident, the presence of social support, whether threats were made, and what is happening in the relationship when violence takes place. Including the role of neighbourhood characteristics such as the concentration of female headed households or levels of community violence would also be informative. This would provide the framework within which these behaviours take place, and allow for more advanced analysis, deepening our understanding of the causes, so that we may, in the future, craft more effective measures of prevention and abatement for our population.
The author acknowledges the helpful comments of Prof Patricia Anderson, and Dr. Garth Lipps who reviewed this article in various forms and stages of development.
Abramsky, T., C.Watts, C. Garcia-Moreno, K.C Devries, L. Kiss, M. Ellsberg, H. Jansen, L. Heise. 2011. What factors are associated with recent IPV? Findings from the WHO multi-country study on women's health and domestic violence. Biomedical Central Public Health 11:10
Alhabib, S., U. Nur, & R. Jones. 2010. Domestic violence against women: Systematic review of prevalence studies. Journal of Family Violence 25: 369-382.
Amnesty International. 2007. Proposal to parliament to include the highest standards of international criminal law into national legislation to stop violence against women, http://www.amnesty.org/en/library/ asset/AMR38/001/2007/en/f46e6978-d3b9-l Idd-a329-2f46302a 8cc6/amr380012007en.pdf
Anderson, P. Y. 2012. Measuring masculinity in an Afro-Caribbean context. Social and Economic Studies 61(1): 49-93.
Bachman, R., & L.E. Saltzman. 1995. Violence against women: Estimates from the redesigned survey. Washington, DC: US Department of Justice, Office of Justice Programs, Bureau of Justice Statistics.
Barrow, C. 2009. "Contradictory sexualities: From vulnerability to empowerment for adolescent girls in Barbados." In Sexuality, social exclusion and human rights: Vulnerability in the Caribbean context of HIV, edited by C. Barrow, M. de Bruin, & R. Carr, 215-38. Kingston, Jamaica: Ian Randle.
Bates, L., S.R. Schuler, F. Islam, & M.K. Islam. 2004. Socioeconomic factors and processes associated with domestic violence in rural Bangladesh. International Family Planning Perspectives 30 (4): 190-199.
Belsky, J. 1980. Child maltreatment: An ecological integration. American Psychologist 35: 320-335.
Bott, S., A. Guedes, M. Goodwin & J.A. Mendoza. 2012. Violence Against Women in Latin America and the Caribbean: A comparative analysis of population-based data from twelve countries. Washington, DC: Pan American Health Organization.
Brokaw, J., L. Fullerton-Gleason, L. Olson, C. Crandall, S. McLaughlin, & D. Sklar. 2002. Health status and intimate partner violence: a cross-sectional study. Annals of Emergency Medicine 39(1): 31-38.
Brown, J., P. Anderson & B. Chevannes. 1993. The contribution of Caribbean men to the family. Report of a Jamaican Pilot Study submitted to 1DRC by the Caribbean Child Development Centre--ARCHIV 94353. http:// idl bnc.idrc.ca/dspace/bitstream/10625/11659/1/94353.pdf
Brownridge, D. A. 2004. Understanding Women's Heightened Risk of Violence in Common-Law Unions Revisiting the Selection and Relationship Hypotheses. Violence Against Women 10 (6): 626-651.
Carlson, B.E. 1984. Causes and maintenance of domestic violence: An ecological analysis. Social Service Review 58: 569-587.
Chevannes, B. 2001. Learning to be a man: Culture, socialization and gender identity in five Caribbean communities. Kingston, Jamaica: The University of the West Indies Press.
Chevannes, B., & C. Mitchell-Kernan. 1992. How we were grown: Cultural aspects of high-risk sexual behaviour in jamaica, Mona. AIDS Project on Sexual decision-making among Jamaicans: USAID & FHI
Dahlberg, L.L., & E.G. Krug. 2002. Violence: A global public health problem. In World report on violence and health, edited by E. Krug, L.L. Dahlberg, J.A. Mercy, A,B, Zwi, & R. Lozano, 1-56 . Geneva: World Health Organization.
DeShong, H. 2011. Gender, Sexuality and Sexual Violence: A Feminist Analysis of Vincentian Women's Experiences in Violent Heterosexual Relationships. University of the West Indies, Cave Hill Campus, Institute for Gender and Development Studies: Nita Barrow Unit (IGDS: NBU) Seminar Paper.
DeKeseredy, W. S., & M. D. Schwartz. 2005. "Masculinities and interpersonal violence." In Handbook of studies on men and masculinities, edited by M. S. Kimmel, J. Hearn, & R.W. Connell, 353-366. Thousand Oaks, CA: Sage.
Dobash, R. E., & R. P. Dobash. 1997. Violence against wives and women, gender violence: Interdisciplinary perspectives. New York, NY: New York University Press.
Doe, S. S. 2000. Cultural factors in child maltreatment and domestic violence in Korea. Children and Youth Services Review 22(3): 231-36.
Dutton, M.A. 1996. "Battered women's strategic response to violence: The role of context." In Future interventions ivith battered women and their families, edited by J. Edleson & Z. Eiskovitz, 105-124. Thousand Oaks, CA: Sage.
Ellsberg, M. C., R. Pena, A. Herrera, J. Liljestrand, & A. Winkvist. 2000. Candies in hell: Women's experiences of violence in Nicaragua. Social Science and Medicine 51: 1595-1610.
--. 1999. Wife abuse among children of childbearing age in Nicaragua. American Journal of Public Health 89: 241-244.
Fagan, J., & A. Browne. 1994. "Violence between spouses and intimates: Physical aggression between men and women in intimate relationships." In Understanding and preventing violence: Vol. 3. Social influences, edited by A. Reiss & J. Roth, 115-292. Washington, DC: National Academy.
Felson, R. B., & S.F. Messner. 2000. The control motive in intimate partner violence. Social Psychology Quarterly 63(1): 86-94.
Gage, A. J. 2005. Women's experience of intimate partner violence in Haiti. Social Science and Medicine 61: 343-364.
Gage, A. J., & P. L. Hutchinson. 2006. Power, control, and intimate partner sexual violence in Haiti. Archives of Sexual Behavior 35:11-24.
Garcia-Moreno C., H.A.F.M. Jansen, M. Ellsberg, L. Heise & C.H.Watts. 2006. The prevalence of intimate partner violence: Findings from the WHO multi-country study on women's health and domestic violence. Lancet 368:1260-69.
Gelles, R. J. 1974. The violent home: A study of physical aggression between husbands and wives. Beverly Hills, CA: Sage.
Gelles, R. J. and M.A. Straus. 1998. Intimate violence: the causes and consequences of abuse in the American family. New York: Simon and Schuster.
Gibbison, G. A. 2007. Attitude towards intimate partner violence against women and risky sexual choices of Jamaican males. West Indian Med J 56(1): 66-71.
Gilligan, J. 2001. Preventing violence. London: Thames & Hudson.
Goetz, A. T., T.K. Shackelford, V.G. Starratt, & W.F. McKibbin. 2008. Intimate partner violence. In Evolutionary forensic psychology, edited by J. D. Duntley & T. K. Shackelford: 65-78. New York: Oxford University Press.
Gopaul, R., & M. Cain. 1996. Violence between spouses in Trinidad and Tobago: A research note. Caribbean Quarterly 42(2-3): 29-41.
Haniff, N. 1998. Male violence against women and men in the Caribbean: The case of Jamaica. Journal of Comparative Family Studies; Summer 1998; 29, 2.
Haniff, N. 1995. Male violence against women and men in the Caribbean: The case of Jamaica. Bridgetown, Barbados: Women and Development Unit.
Harriot, A. ed. 1996. Understanding crime in Jamaica: New challenges for public policy. Kingston, Jamaica: University of the West Indies Press.
Hearn, Jeff. 2012. The sociological significance of domestic violence: Tensions, paradoxes and implications. Current Sociology :1-19. doi: 10.1177/0011392112456503
Heise, L.L. 1998. Violence against women: An integrated, ecological framework. Violence Against Women 4(3): 262-290.
Hindin, M., & L.S. Adair. 2002. Who's at risk? Factors associated with intimate partner violence in the Philippines. Social Science and Medicine 55:1385-1399.
Hoffman, K., D. H. Demo & J. N. Edwards. 1994. Physical wife abuse in a non-Western society: An integrated theoretical approach. J Marriage Fam. 56: 131-146.
Jewkes, R. 2002. Intimate partner violence: Causes and prevention. Lancet 359 (9315): 1423-1429.
Jewkes, R., & N. Abrahams. 2002. The epidemiology of rape and sexual coercion in South Africa: An overview. Social Science and Medicine 55(7): 1231-1244.
Jewkes, R., K. Dunkleb, M. P. Kossc, J. B. Levind, M. Ndunae, J. Nwabisa & Y. Sikweyiya. 2006. Rape perpetration by young, rural South African men: prevalence, patterns and risk factors. Social Science and Medicine 63(11): 2949-2961.
Jewkes, R., J. Levin & L. Penn-Kekana. 2002. Risk factors for domestic violence: Findings from a South African cross-sectional study. Social Science and Medicine 55(9): 1603-1617.
Johnson, M P. November 2000. Conflict and control: Symmetry and asymmetry in domestic violence. Paper presented at the National Institute of Justice Gender Symmetry Workshop, Arlington, VA.
Jones, M. 2003. "From the footnotes and into the text: Victimization of Jamaican women." In Understanding crime in Jamaica: New challenges for public policy, edited by A. Harriot, 113-132. Kingston, Jamaica: University of the West Indies Press.
Kempadoo, K. 2009. Caribbean sexuality: Mapping the field. Caribbean Review of Gender Studies 3: 1-24.
Krug, E., J. Mercy, L. Dahlberg & A. Zwi. 2002. Intimate partner violence: Causes and prevention. Lancet 360: 1083-1088.
Lau, U. 2009. Intimate partner violence. (Fact sheet). Cape Town: MRC-UNISA Crime, Violence and Injury Lead Programme. http://www. mrc.ac.za/crime/intimatepartner.pdf
Le Franc, E., & L. Rock. 2001. The commonality of gender based violence. Journal of Eastern Caribbean Studies 26(1): 74-82.
Le Franc, E., M. Samms-Vaughan, I. Hambleton, K. Fox, & D. Brown. 2008. Interpersonal violence in three Caribbean countries: Barbados, Jamaica, and Trinidad and Tobago. Pan American Journal of Public Health 24(6): 409-421.
Le Franc, E., G. E. Wyatt, C. Chambers, D. Eidemire, B. Bain & H. Ricketts. 1996. Working women's sexual risk taking in Jamaica. Social Science and Medicine 42(10): 1411-1417.
Levendosky, A., B. Lannert & M. Yalch. 2012. The effects of intimate partner violence on women and child survivors: an attachment perspective. Psychodynamic Psychiatry 40 (3): 397-434.
Levinson, D. 1989. Family violence in cross-cultural perspective. Thousand Oaks, CA: Sage.
Lupri, E. 1990. Harmony and aggression: On the dialectic of spousal violence. Kolner Zeitschrift fur Soziologie und Sozialpsychologie 42(3): 474-501.
Michalski, J. 2004. Making sociological sense out of trends in intimate partner violence: the social structure of violence against women. Violence Against Women 10(6): 652-675.
National Family Planning Board. 2010. Reproductive health survey jamaica 2008 Report. Kingston, Jamaica: National Family Planning Board.
Parsad, B. S. 1988, May. Domestic violence: A study of wife abuse among East Indians of Guyana. Paper presented at Conference on the Genesis of a Nation: Origins and Development of Indo-Guyanese, Georgetown.
Peake, L., & A. D. Trotz. 1999. Gender, ethnicity and place: Women and identities in Guyana, Routledge Studies in Development and Society. New York: Routledge.
Platek, S., & T. Shackelford. 2006. "Introduction to theory and research on anti-cuckoldry tactics: Overview of current volume." In Female infidelity and paternal uncertainty: Evolutionary perspectives on male anti-cuckoldry tactics edited by S. Platek & T. Shackelford, 1-13. Cambridge: Cambridge University Press.
Rao, V. 1997. Wife beating in rural south India: A qualitative and econometric analysis. Social Science and Medicine 44:1169-80.
Romans, S., T. Forte, M. M. Cohen, J. Du Mont, & I. Hyman. 2007. Who is most at risk for intimate partner violence? A Canadian population-based study. Journal of interpersonal violence 22(12): 1495-1514.
Rosenberg, M.L., A. Butchart, J. Mercy, V. Narasimhan, H. Waters, & M. S. Marshall. 2006. Interpersonal violence. In Disease Control Priorities in Developing Countries, edited by D.T. Jamison, J.G. Breman, A. R. Measham, G. Alleyne, M. Claeson, D.G. Evans, J. Prabhat, A. Mill, & P. Musgrove, 2nd edition, 755-770. Washington, DC: Oxford University Press and The World Bank.
Schuler, S. R. 1996. Credit programs, patriarchy and men's violence against women in rural Bangladesh. Social Science and Medicine 43(12): 1729-1742.
Schuler, S.R., S.M. Hashemi & S.H. Badal. 1998. Men's violence against women in rural Bangladesh: Undermined or exacerbated by micro-credit programmes? Development in Practice 8(2): 148-157.
Sorenson, S. B., D.M. Upchurch & S. Haikang. 1996. Violence and injury in marital arguments: Risk patterns and gender differences. American Journal of Public Health 86(1): 35-40.
Straus, M. A., S. L. Hamby, S. Boney-McCoy & D. B. Sugarman. 1996. The revised Conflict Tactics Scales (CTS2) development and preliminary psychometric data. Journal of family issues 17(3): 283-316.
Straus, M. A., R. J. Gelles & S.K. Steinmetz Eds.. 1980. Behind closed doors: Violence in the American family. Piscataway, N.J.: Transaction.
Tjaden, R, & N. Thoennes. 2000. Full report of the prevalence, incidence, and consequences of violence against women: findings from the National Violence Against Women Survey. Washington, DC: United States Department of Justice 71.
UNDP. 2012. Caribbean Human Development Report 2012: Human Development and the shift to better citizen security. New York: UNDP.
UNFPA. 2009. Country baseline assessment of gender-based violence. Report I done in collaboration with the Institute for Gender and Development, University of the West Indies. Kingston, Jamaica: United Nations Population Fund.
United Nations. 1993. Declaration on the elimination of violence against women. General Assembly Resolution Document A/RES/48/104. New York: United Nations.
Wilson, M., & M. Daly. 1992. "Till death do us part." In Femicide: The politics of woman killing, edited by J. Radford & D. E. H. Russell, 83-98. New York: Twayne.
World Health Organization. 2010. Preventing intimate partner and sexual violence against women: Taking action and generating evidence. Geneva: WHO and London School of Hygiene and Tropical Medicine.
Yllo, K, & M. A. Straus. 1981. Interpersonal violence among married and cohabiting couples. Family Relations: 339-347.
(1) A visiting union exists when a man and a woman have a steady sexual relationship but are neither legally married to each other nor living with each other.
(2) These include, for example, emergency room admissions, or health insurance claimants.
(3) This study differs from the Bott et al., 2012 analysis which did not include the acceptance of gender norms and the controlling behaviour of partners together with early childhood exposure and women's background variables, simultaneously in the multivariate models presented.
(4) Exposure to the ridicule of community and peers heightens the threat involved as a man's lack of control over his partner labels him as a failure.
(5) See Brown, Anderson & Chevannes (1993: 65) and Chevannes (1986) for a fuller discussion on the conditions under which Jamaican men accept a "jacket".
(6) Some of these include economic inequality between partners resulting in women's dependence, the widespread acceptance of violence to resolve conflicts, the virtual absence of effective legislation to protect women's rights, and women's ignorance of measures to prevent the occurrence (NFPB 2010: 6)
(7) According to the NFPB, women subjected to domestic violence may be unable to use contraception effectively and consistently, and may lack control or negotiation skills that will enable them to avoid unintended pregnancies and sexually transmitted diseases, including HIV/AIDS (2010: 6).
(8) Although the survey included men (15-24 years) living in households, the household selection for the male sample was independent from the selection of households for the female sample. We describe the selection of the female sample here.
(9) Education refers to the number of years of schooling that the respondent received; with eight years or less being indicative of primary level attained, 9-13 years as secondary level, and 14 or more years reflecting a post-secondary level of attainment. The cut-off of eight years was used since this more accurately represents the number of years spent in primary schooling through to grade six, by those attending public primary schools, preparatory schools, all-age and elementary schools.
(10) Those who were employed were considered to be working, while those unemployed, keeping house, students, or who were incapable of working, were classified as not working. Parity reflects the number of children a woman has borne.
(11) See NFPB (2010: 17) for a discussion of the wealth quintiles.
Table 1: Percent distribution of women who have had at least one partnership in their lifetime and experienced IPV in the past year, by various socio-demographic characteristics Socio-demographic Experienced No experience All characteristics IPV of IPV Women Age 15-24 27.6 24.3 24.9 25-34 41.6 34.5 35.7 35-49 30.7 41.2 39.4 Union status Married 14.8 18.2 17.6 Common-law 37.1 24.8 26.9 Visiting 36.0 37.0 36.8 Not currently 12.2 20.0 18.7 with a partner Years of education completed 0-9 27.5 21.3 22.3 10-12 63.9 63.9 63.9 13 or more 8.7 14.8 13.8 Place of residence Urban 56.9 57.8 57.6 Rural 43.1 42.2 42.4 Work status Working 46.2 49.5 49.0 Employed 46.2 49.5 49.0 Not working 53.7 50.4 51.0 Unemployed 17.4 13.7 14.3 Keeping house 28.3 24.4 25.1 Student 3.8 5.4 5.1 At home, not 3.8 6.4 6.0 keeping house Incapable of working 0.4 0.5 0.5 Household wealth quintile Lowest 29.3 20.5 22.0 Second lowest 23.7 18.3 19.2 Middle 21.7 19.4 19.8 Second highest 14.2 21.3 20.1 Highest 11.1 20.5 18.9 Parity 0 19.6 21.7 21.4 1-2 43.2 25.5 46.2 3 or more 37.2 52.8 32.4 Total 100.0 100.0 100.0 n 8,018 1,624 9,641 Table 2: Number and percent distribution of women who have had at least one partnership in their lifetime, by selected health characteristics Health Characteristic Experienced No experience All IPV of IPV Women Currently pregnant 5.3 4.7 4.8 Not currently pregnant 94.7 95.3 95.2 Total 100.0 100.0 100.0 n 1,605 7,977 9,582 Used contraception 74.2 74.9 74.8 at last sex Did not use contraception 25.8 25.1 25.2 at last sex Total 100.0 100.0 100.0 n 1,580 6,983 8,563 Age at first sex Under 15 21.1 12.1 13.7 15-19 72.8 74.2 73.9 20 or older 6.1 13.7 12.4 Total 100.0 100.0 100.0 n 1,545 7,586 9,131 Self-rated health status Excellent 21.0 27.2 26.1 Very good 40.9 47.5 46.4 Fair 32.0 21.7 23.4 Poor 6.1 3.7 4.1 Total 100.0 100.0 100.0 n 1,608 7,986 9,594 Table 3: Number and percent distribution of women who witnessed violence between their parents, before the age of 15 years Witnessed her father/ Experienced No experience All stepfather hit, slap, IPV of IPV Women lick, shove her mother/ stepmother before age 15 Yes 30.9 17.3 18.0 No 69.1 82.7 73.8 Total 100.0 100.0 100.0 n 1,526 7,326 8,852 Note: Respondents who could not recall or who did not grow up with both parents were removed from the tabulations. Table 4: Number and percent distribution of women who were physically hurt by someone over the course of their lifetime Characteristic Experienced No experience All IPV of IPV Women Physically hurt by parent or own family member before age 15 Parent or other 81.0 68.5 70.2 adult family member Was not hurt 19.0 31.5 29.8 Total 100.0 100.0 100.0 n 1,622 7,965 9,587 Physically hurt by someone other than a current or past partner after age 15 Yes 45.4 23.1 26.7 No 54.6 76.9 73.3 Total 100.0 100.0 100.0 n 1,616 7,973 9,589 Table 5: Number and percent distribution of women expressing agreement with statements reflecting the controlling behaviour of her current partner It is true that her Experienced No All current partner: IPV experience Women n = 1,619 of IPV n = 7,992 n = 9,611 Tries to keep her from 19.2 4.9 7.3 seeing her friends and family Experienced No All IPV experience Women n = 1,617 of IPV n = 9,603 n = 7,986 Insists on knowing where 57.2 30.5 34.8 she is at all times Experienced No All IPV experience Women n = 1,618 of IPV n = 9,600 n = 7,982 Gets angry if she speaks 60.1 24.2 30.1 with another man Experienced No All IPV experience Women n = 1,622 of IPV n = 9,603 n = 7,981 Is often suspicious that 45.4 14.9 20.0 she is unfaithful Experienced No All IPV experience Women n = 1,614 of IPV n = 9,582 Expects her to ask his 9.0 3.3 4.2 permission for healthcare Table 6: Number and percent distribution of women expressing agreement with statements indicating reasons why a man may have good reason to hit his wife A man has a good reason Experienced No All to hit his wife if: IPV experience Women n = 1,622 of IPV n = 9,631 n = 8,009 She does not complete her 1.2 0.6 0.7 household work to his satisfaction Experienced No All IPV experience Women n = 1,617 of IPV n = 9,621 n = 8,004 She disobeys him 1.1 0.8 0.8 Experienced No All IPV experience Women n = 1,619 of IPV n = 9,623 n = 8,004 She refuses to have 0.4 0.2 0.3 sexual relations with him Experienced No All IPV experience Women n = 1,617 of IPV n = 9,621 n = 8,004 She asks him whether he 1.1 0.8 0.3 has other girlfriends Experienced No All IPV experience Women n = 1,603 of IPV n = 9,581 n = 7,978 He finds out that she has 3.0 2.3 2.4 been unfaithful Table 7: Number and percent distribution of women expressing agreement with statements reflecting acceptance of gender norms Norm Experienced No All IPV experience Women n = 1,591 of IPV n = 9,396 n = 7,805 A good wife obeys her 46.1 51.4 50.5 husband even if she disagrees Experienced No All IPV experience Women n = 1,610 of IPV n = 9,534 n = 7,921 Family problems should 65.6 64.8 65.0 only be discussed with people in the family Experienced No All IPV experience Women n = 1,605 of IPV n = 9,506 n = 7,901 It is important for a man 22.2 16.5 17.5 to show his wife/partner who is boss Experienced No All IPV experience Women n = 1,597 of IPV n = 9,500 n = 7,903 It is a wife's obligation 26.4 31.7 30.8 to have sex with her husband even if she doesn't feel like it Experienced No All IPV experience Women n = 1,582 of IPV n = 9,432 n = 7,850 It is alright for a wife 71.0 67.5 68.1 to refuse sex with her husband if he sees other women Experienced No All IPV experience Women n = 1,555 of IPV n = 9,254 n = 7,699 If a man mistreats his 68.2 63.5 64.3 wife, others outside the family should intervene Table 8: Number and percent of women experiencing individual acts of violence perpetrated by their male partners in the last 12 months, by quintile Wealth Quintile I II In the last 12 months her partner has: No. % No. % Physical Slapped her or threw 144 6.8 105 5.7 something at her Pushed, shoved her, or 112 5.3 78 4.2 pulled her hair Hit her with his fist or 135 6.4 101 5.5 something harmful Threatened her with a 42 2.0 33 1.8 gun, knife or other weapon Choked or burnt her on 42 2.0 33 1.8 purpose Kicked, dragged or beat 72 3.4 33 1.8 her up Experienced any physical 183 8.6 171 9.2 act of violence Sexual Physically forced to have 89 4.2 75 4.1 intercourse when unwanted Experienced any verbal, physical, or sexually 476 22.4 384 20.8 violent act Total number of 2,122 100.0 1,849 100.0 respondents Wealth Quintile III IV V In the last 12 months her partner has: No. % No. % No. Physical Slapped her or threw 94 4.9 59 3.0 30 something at her Pushed, shoved her, or 80 4.2 40 2.1 37 pulled her hair Hit her with his fist or 86 4.5 53 2.7 29 something harmful Threatened her with a 30 1.6 16 0.8 16 gun, knife or other weapon Choked or burnt her on 41 2.1 15 0.7 9 purpose Kicked, dragged or beat 58 3.0 25 1.3 14 her up Experienced any physical 133 7.0 83 4.3 57 act of violence Sexual Physically forced to have 45 2.4 38 2.0 26 intercourse when unwanted Experienced any verbal, physical, or sexually 353 18.5 230 11.9 181 violent act Total number of 1,912 100.0 1,938 100.0 1,820 respondents Wealth Quintile V All In the last 12 months her partner has: % No. % Physical Slapped her or threw 1.6 432 4.5 something at her Pushed, shoved her, or 2.0 347 3.6 pulled her hair Hit her with his fist or 1.6 404 4.2 something harmful Threatened her with a 0.9 137 1.4 gun, knife or other weapon Choked or burnt her on 0.5 140 1.5 purpose Kicked, dragged or beat 0.8 202 2.1 her up Experienced any physical 3.1 627 6.5 act of violence Sexual Physically forced to have 1.4 273 2.8 intercourse when unwanted Experienced any verbal, physical, or sexually 9.9 1,624 16.8 violent act Total number of 100.0 9,641 100.0 respondents Note: Percents are calculated as a proportion of the total in the respective quintile category. Percents will not add to 100 because the acts are not mutually exclusive. Table 9: Average number of violent acts experienced per woman, by wealth quintile Type of violent act experienced Wealth quintile I II III IV V All Any violent act 1.21 1.14 1.19 0.89 0.76 1.08 Physical 1.10 0.91 1.27 0.84 0.76 1.01 Sexual 0.42 0.44 0.35 0.32 0.25 0.37 Note: Average is based on the total in each quintile category. Table 10: Results of logistic regression analysis for characteristics predicting exposure to any form, physical and sexual violence between intimate partners Variables Any form Odds SE Sig. Ratio SOCIO-DEMOGRAPHIC Age Group (35 years or older=ref) Age 15-24 1.833 .105 .000 Age 25-34 1.843 .081 .000 Education (Post- secondary=ref) Primary 1.240 .184 .243 Secondary 1.131 .145 .395 Union status (Married=ref) Common-law .918 .104 .411 Visiting .563 .106 .000 No steady partner .364 .124 .000 Parity (3 or higher=ref) Childless 1.179 .112 .140 1-2 children .990 .082 .900 Health status (rated good or better=ref) Fair 1.398 .072 .000 Poor 1.509 .145 .005 Number of lifetime partners (One=ref) Two 1.410 .085 .000 Three or more 1.570 .086 .000 Work status (Not working=ref) Working 1.095 .068 .181 HOUSEHOLD Wealth quintile (V=ref) I--lowest 2.280 .126 .000 II--second 1.909 .121 .000 III--middle 1.655 .117 .000 IV--next 1.185 .121 .159 Place of residence (Rural=ref) Urban 1.201 .070 .009 SOCIALISED TOWARDS VIOLENCE Witnessed violence between parents in childhood (no=ref) Experienced it 1.289 .074 .001 Experienced violence from a parent or other adult in family in childhood (no=ref) Witnessed it 1.406 .081 .000 Experienced violence from non-partner after age 15 (no=ref) Experienced it 1.732 .069 .000 GENDER NORMS Conformity to gender norms (agrees with less than two statements = ref) Agrees with 2-4 .708 .072 .000 statements Agrees with 5-6 .622 .096 .000 statements PARTNER CONTROLLING BEHAVIOUR Partner exercises no control (agrees with no statement = ref) Agrees with 1-2 2.873 .078 .000 statements describing partner control Agrees with 3-6 8.036 .084 .000 statements describing partner control Constant .021 .202 .000 n = 6,368 -2LL (intercept) 7965.583 -2LL (final) 6519.047 [Model.sup.2] 1524.467 Degrees of freedom 26 P = .000 Nagelkerke [R.sup.2] .266 Variables Physical Odds SE Sig. Ratio SOCIO-DEMOGRAPHIC Age Group (35 years or older=ref) Age 15-24 3.007 .160 .000 Age 25-34 2.097 .128 .000 Education (Post- secondary=ref) Primary 1.414 .295 .241 Secondary 1.166 .249 .538 Union status (Married=ref) Common-law 1.598 .175 .007 Visiting .836 .180 .319 No steady partner .727 .200 .110 Parity (3 or higher=ref) Childless .785 .169 .151 1-2 children .864 .124 .240 Health status (rated good or better=ref) Fair 1.104 .108 .358 Poor 1.392 .202 .101 Number of lifetime partners (One=ref) Two 1.173 .128 .211 Three or more .919 .131 .517 Work status (Not working=ref) Working .925 .102 .444 HOUSEHOLD Wealth quintile (V=ref) I--lowest 2.323 .199 .000 II--second 2.153 .189 .000 III--middle 1.654 .189 .008 IV--next 1.382 .198 .103 Place of residence (Rural=ref) Urban 1.974 .108 .000 SOCIALISED TOWARDS VIOLENCE Witnessed violence between parents in childhood (no=ref) Experienced it 1.195 .106 .093 Experienced violence from a parent or other adult in family in childhood (no=ref) Witnessed it 1.176 .126 .196 Experienced violence from non-partner after age 15 (no=ref) Experienced it 2.900 .101 .000 GENDER NORMS Conformity to gender norms (agrees with less than two statements = ref) Agrees with 2-4 .589 .106 .000 statements Agrees with 5-6 .636 .140 .001 statements PARTNER CONTROLLING BEHAVIOUR Partner exercises no control (agrees with no statement = ref) Agrees with 1-2 2.448 .138 .000 statements describing partner control Agrees with 3-6 10.135 .132 .000 statements describing partner control Constant .003 .345 .000 n = 6,368 -2LL (intercept) 4277.518 -2LL (final) 3282.814 [Model.sup.2] 994.703 Degrees of freedom 26 P = .000 Nagelkerke [R.sup.2] .278 Variables Sexual Odds SE Sig. Ratio SOCIO-DEMOGRAPHIC Age Group (35 years or older=ref) Age 15-24 1.762 .218 .009 Age 25-34 1.502 .177 .021 Education (Post- secondary=ref) Primary 2.129 .384 .049 Secondary .935 .336 .841 Union status (Married=ref) Common-law .543 .239 .011 Visiting .581 .232 .019 No steady partner .614 .253 .054 Parity (3 or higher=ref) Childless 1.976 .227 .003 1-2 children 1.697 .178 .003 Health status (rated good or better=ref) Fair 1.275 .149 .102 Poor 1.567 .259 .083 Number of lifetime partners (One=ref) Two 1.785 .200 .004 Three or more 2.280 .196 .000 Work status (Not working=ref) Working .722 .147 .027 HOUSEHOLD Wealth quintile (V=ref) I--lowest 2.491 .281 .001 II--second 2.005 .271 .010 III--middle 1.120 .280 .685 IV--next 1.347 .278 .284 Place of residence (Rural=ref) Urban 1.775 .151 .000 SOCIALISED TOWARDS VIOLENCE Witnessed violence between parents in childhood (no=ref) Experienced it 1.293 .149 .085 Experienced violence from a parent or other adult in family in childhood (no=ref) Witnessed it .749 .159 .070 Experienced violence from non-partner after age 15 (no=ref) Experienced it 1.325 .145 .052 GENDER NORMS Conformity to gender norms (agrees with less than two statements = ref) Agrees with 2-4 1.006 .156 .967 statements Agrees with 5-6 1.023 .196 .909 statements PARTNER CONTROLLING BEHAVIOUR Partner exercises no control (agrees with no statement = ref) Agrees with 1-2 2.577 .216 .000 statements describing partner control Agrees with 3-6 11.952 .197 .000 statements describing partner control Constant .002 .476 .000 n = 6,368 -2LL (intercept) 2337.547 -2LL (final) 1899.016 [Model.sup.2] 438.531 Degrees of freedom 26 P = .000 Nagelkerke [R.sup.2] .208
|Printer friendly Cite/link Email Feedback|
|Publication:||Social and Economic Studies|
|Date:||Mar 1, 2014|
|Previous Article:||Avoiding vertical inefficiencies in funding tertiary level education (TLE) in resource abundant states/Evitar las ineficacias verticales en la...|
|Next Article:||Coral gardens 1963: the Rastafari and Jamaican independence.|