Introduction

A lack of gender equality is consistently cited as an underlying determinant of violence against women. The United Nations General Assembly, in its 1993 Declaration on the Elimination of Violence Against Women, noted that this violence is a manifestation of historically unequal power relations between men and women. Gender inequality as a cause of violence against women also underpins approaches to prevention by organisations such as the World Health Organization (World Health Organization (WHO), 2010), and, in Australia, VicHealth (VicHealth, 2007), as well as much of the research on the topic (Yodanis, 2004). So it has long been understood that gender inequality is a problem, yet the evidence of the link between gender inequality and violence against women is rarely laid out clearly to illustrate this connection. With whole frameworks based on the premise that prevention of violence against women requires improved gender equality, it is important to summarise and examine the evidence on this correlation.

Achieving gender equality is a key goal in the prevention of violence against women by those aiming to reduce gendered violence (WHO, 2005, 2010; VicHealth, 2007). However, without an existing model of gender equality to demonstrate the end product, and a lack of research and data around whether some aspects of gender equality are more important than others in preventing violence, it is difficult to pin down the impact that gender inequality actually has as a determinant of violence against women, despite the obvious logic in the connection. It appears that the vital element to consider is the gender norms and beliefs surrounding male dominance and male superiority, created by power hierarchies that accord men greater status. However, acknowledging social norms as a key factor means that issues of measurement and definition become more complex. How then do we determine exactly when we have reached or come closer to a position of gender equality? How do we incorporate or measure other sources of disadvantage that compound inequality?

A key focus of this paper will be to trace the path of research on violence against women to the conclusion that gender inequality is an underlying determinant in the factors that cause it. This research summary will look at the relationship between gender inequality and gendered violence and the evidence that is out there to support this connection. In summarising the literature about gender equality, this paper will consider what is actually meant by gender equality, how it is defined and how it is measured.

What does become clearer in summarising the research on gender equality and violence against women is the importance of using an ecological model of prevention that enables influencing factors at various levels - societal, community and individual - to give greater contextual meaning to how gender inequality plays out in reality. This paper focuses on the societal factors that link to violence against women, rather than community and individual factors that are more easily identified within particular settings.

The acknowledgement of gender inequality as an underlying determinant of violence against women has seen the implementation of gender equality become a strategic policy goal in many countries and international organisations such as the World Health Organization. This is sometimes termed "gender mainstreaming" - a process of updating key concepts to enable an understanding of the world through a gendered lens (Walby, 2005). Though the benefits of this approach are quite obvious in acknowledging the effects of gender disadvantage, there has also been some criticism that it fails to acknowledge other sources of and intersections of disadvantage such as class and race, which may compound gender disadvantage.

Some conceptual definitions

Gender

The concept of gender is used here in reference to the social configuration of men and women's identities, interactions and relationships rather than the biological differences of sex that are male and female (Victorian Health Promotion Foundation, Flood, & Pease, 2006). Gender systems refer to the socio-cultural environments whereby expectations of and values for each gender are determined within the cultural context (UN Office of the Special Advisor to the Secretary-General on Gender Issues and the Advancement of Women [OSAGI], n.d.). These gender systems are constructed and maintained through institutional and political structures such as policy, legislation and economic requirements. Gender roles are then learned through the process of socialisation and cultural expectations (OSAGI, n.d.).

Gender equality

The United Nations, UNICEF and the World Health Organization, among others, use gender equality as their preferred terminology. The term "gender equality" does not mean that men and women are necessarily exactly the same or that differences don't exist, but that they have equal rights, opportunities, responsibilities and access to resources as well as the enjoyment of them. It means that differences in roles and biology, that may impact on equal enjoyment of these rights and responsibilities are taken into account by ensuring that men and women's perceptions, interests, needs and priorities are given equal weight (OSAGI, n.d.). Therefore, the term gender equality (or inequality) will be referred to throughout this paper as the issue under examination when considering underlying determinants of violence against women.

Gender equity

A search of the term "gender equity" indicates it is less clear in meaning than the term gender equality and is sometimes used interchangeably and sometimes only in the context of economics, for example in gender pay equity. United Nations' definitions indicate that gender equity is about interpreting policy and political and other social processes within a frame of historical disadvantage for women, to ensure fairness (UNFPA United Nations Population Fund, 2008). Equitable strategies must be able to accommodate differences between men and women and balance out the disadvantages that women have historically suffered. This enables a basis for women and men to equally enjoy social benefits, for example in terms of opportunities, economic benefits, power and decision-making and resources. Therefore, gender equity will be the pathway to gender equality (UNFPA United Nations Population Fund, 2008; OSAGI, n.d.).

Determinant

In a public health sense or prevention sense, a determinant is a "foundational" cause of a particular health issue or social problem (Hankivsky & Christoffersen, 2008). Determinants can include the social conditions in which people live that impact and shape their experience of a health or social issue. They can be referred to as the causes behind the cause (Quadara, Nagy, Higgins, & Siegal, 2014), or the set of underlying conditions that enable the risk factors for experiencing a particular health issue such as violence against women.

How did we get here? Establishing gender inequality as a determinant for violence against women

It has been acknowledged that there is a lack of information about causal factors for violence against women at the societal level of the ecological model compared to the individual and community levels (Krug, Dahlberg, Mercy, Zwi, & Lozano, 2002; WHO, 2010). Indeed, the lack of research data about societal risk factors makes comparison between settings within countries or between countries difficult to explore.

Cross-cultural studies provide important clues to understanding the ways in which social norms in different cultures affect levels of gendered violence. The work that has been done in this area finds a correlation between cultural social roles and levels of violence (Archer, 2006).

However, establishing the underlying causes of violence against women has been the subject of varying theories. The most prominent of these was the feminist model of causation that came out of the feminist movement of the 1970s. This held that patriarchy and men's indisputable power and oppression of women were the underlying causes (WHO, 2010). More recently, extended models of explanation have been incorporated to develop the theory of gender inequality further and to explore the effects of social roles, attitudes and other cultural factors. The one currently most in use is the ecological model. Heise (1998) advocated for a conceptualisation in an ecological model that could take feminist theory further, incorporating other societal and community factors influencing individual perpetrator behaviour and explaining why some but not all men perpetrate violence.

The work of feminist activists in raising the profile of intimate partner violence and other violence against women enabled the issue to be seen as a significant public health problem. This has facilitated an approach that aims for prevention of the problem. However, one of the key factors in public health prevention is identifying the societal factors of the problem at hand and working to change these. This is more difficult where the problem is not just physical or biological but has a social dimension as well, as is the case with violence against women. This requires investigation of social elements at various levels of the ecological model.

The role of social norms and gendered expectations

An important aspect of thinking about gender inequality in this societal sense is to understand the role of social norms and social organisation in situating groups into a hierarchical structure based on certain features such as gender, race or class (Ridgeway, 2014).

Such structures result in social status being conferred upon those with the most advantageous position. In terms of power and access to resources, these structures sort groups into a perpetuating pattern of inequality as the higher status groups retain their hold on power and resources. Ridgeway observed that social status or status in relation to other groups is therefore a central consideration in entrenched inequality. Entrenched differences in status lead to "status beliefs" and expectations about the social difference, for example that males are more competent, that then become autonomous beliefs which continue to reproduce the differences by perpetuating patterns of inequality (Ridgeway, 2014). This conceptualisation of status explains the widely shared cultural status beliefs at the societal level that impact on ordinary social relations at the community, organisational and individual levels, and work to legitimate the inequality of the social structure (Ridgeway, 2014). An important aspect of viewing inequality as a relational aspect between social group classification is that it can bring in additional components of social inequality such as race, disability and class.

Summarising the evidence around gender inequality as a determinant for violence against women, one of the key features of sexual and family violence is the reality that such violence is disproportionately experienced by women as victims and perpetrated by men (WHO, 2005).

Socio-cultural theories about causes of violence are based on the consideration of power and its relation to social structures such as institutions, political and economic systems as well as shared beliefs and attitudes that may be influenced by these structures. Such shared attitudes and behaviours can be considered to be cultural factors that underlie behaviour in a particular group (Nayak, Byrne, Martin, & Abraham, 2003). In assessing causation of violence against women, interventions that address these specific factors to achieve attitudinal and behaviour change will be most relevant. It is therefore important to look at specific differences cross-culturally and take into account influences such as religion, history and political factors in assessing differences (Nayak et al., 2003). By examining the attitudes across diverse countries and cultures, it is more likely that themes relating to causation of violence against women can be gleaned. It has already been established by research that gender is important in relation to attitudes, for example gender differences in attitudes towards sexual assault and domestic violence have been noted between men and women, with men more likely to endorse violence against women (Fanslow, Robinson, Crengle, & Perese, 2010; Nayak et al., 2003; Victorian Health Promotion Foundation et al., 2006).

Many current models of primary prevention use the ecological model developed by Bronfenbrenner as a conceptual frame by which to understand causes and risk factors of public health issues such as violence against women (WHO, 2010).¹ This approach to gender-based violence enables an understanding of multiple causality and influences at different levels, including the individual, interpersonal or relationship level, community level and societal level. The societal level relates to larger, structural factors such as cultural beliefs and societal norms that create or sustain gaps or tensions between groups of people. It is at this level that gender inequality influences social roles and impacts most profoundly to create structural power differences that filter down to the other levels to create power and resource imbalances that affect gender interactions at the community, relationship and individual levels (WHO, 2010).

Gendered social scripts (that prescribe how each gender should behave) and inequality between the genders are two prominent factors that feature in the evidence base around violence against women (WHO, 2010). They are influences from the societal level of the ecological model but there is an acknowledged lack of research around them and their link to violence against women (WHO, 2010). The power structures also maximise male access to social and political power at the expense of female contributions to decision-making. With this background of social inequality, power differences in particular relationships between men and women and the use of violence as a way of resolving conflicts have been identified as more specific risk factors for gendered violence (WHO, 2010).

In her early work on the ecological framework of violence, Heise (1998) argued for the adoption of this framework with a particular emphasis on the ability of societal factors such as the socialisation of masculinity to be linked to aggression and dominance. She argued that various lines of research supported the suggestion that adherence to rigid gender roles correlates highly with interpersonal violence. Heise's arguments support the need for a more complex and nuanced analysis that is necessary for an understanding of individual and community variations in violence against women. The ecological framework provides a strong basis for a public health approach to prevention of violence against women by enabling theoretical consideration of the diversity of factors that influence inequality at the broadest level.

A public health approach to considering gender inequality and violence against women

Although the types of violence against women cross over and are linked, having similar underlying determinants, there will be differences in the causes of perpetration of intimate partner violence and sexual violence perpetrated outside of an intimate relationship. Therefore, the following information should be considered together and separately when analysing the evidence of underlying drivers of violence against women.

Gender inequality and intimate partner violence

In a review of studies, including cross-cultural studies, Jewkes (2002), the prominent violence against women researcher, considered various theories of causation and risk factors for intimate partner violence against women. Issues such as poverty, power and gender roles were all considered in this review. It was concluded that two factors seem to be necessary for intimate partner violence against women. First is the unequal position of women within particular relationships, and second is the normative use of violence in a conflict. These factors then interact and are enhanced by societal norms about gender roles and ideologies of male superiority and legitimise demonstrations of power against women. Jewkes also noted that the effects of factors contributing at other levels of the social ecology, such as alcohol and women's poverty, are interconnected and dynamic.

Jewkes (2002) noted that where there were higher levels of female empowerment, this seemed to have a protective effect against intimate partner violence. However, the sources of this female power varied and had different levels of effect. For example, education or income may be sources of power for women but may have different impacts depending on cultural values or need. Education was consistently noted as conferring social empowerment for both men and women but the relationship between education and women has more complex effects as it can increase the risk factor for violence. Showing an inverted U-shape, education only offers a protective element at the highest and lowest levels. This means that a low level of education can be protective, a very high level of education can be protective but there is potentially a backlash effect with a moderate level of education (Jewkes, 2002).

It is necessary to compare and contrast differences and similarities between the factors relating to gender equality that feature regularly in cultures with higher or lower rates of violence against women. Jewkes also noted that societies with stronger belief systems about male dominance have more intimate partner violence. These beliefs are also more likely to be reflected at a societal level such as in political and economic systems.

The earlier work of researcher David Levinson (1989) is widely cited as providing a cross-cultural observation of physical violence against women. It looked at different aspects of female status and its relationship with physical violence against women. His ethnographic studies encompassed 90 societies and analysed cultural features that might be related to "wife-beating" (as it was termed in his work). Levinson noted that sexual inequality can be represented and measured differently across different societies and that sometimes there were differences in women's status in private compared to public spheres. His work considered differences in status and power as multidimensional. Status and power were defined and measured in relation to the control of wealth and property, dominance in the household, restrictions on sex and marriage, separation of the sexes, the value of female human life and male dominance in kinship relationships. Levinson's studies showed that "wife-beating" was likely to occur more frequently in societies where men have control over finances and domestic decision-making, divorce is difficult for women to obtain, the husband's kin group controls the wife's ability to remarry, polygamous marriage is permitted and women do not band together in exclusively female workgroups. Levinson's work pointed out that the relationship between physical family violence and sexual equality is very complex but that the presence of all-female work groups provides women with social support and economic independence from men.

Gender inequality and sexual violence against women

The World Association for Sexual Health has noted that even though gender-related behaviour, including sexual behaviour is influenced by culture and social role expectations, despite cultural diversity across the globe, heterosexual sexual activity is notably consistent in the gendered power imbalances that prevail (World Association for Sexual Health, 2008).

In order to study the drivers of perpetration of rape, it is necessary to study men as well as women in relation to this gendered crime (Jewkes, 2012). In the introduction to this review the author noted that there has been very little research done in the form of population studies, and for the purposes of prevention, cross-sectional and longitudinal studies are really required to measure environmental and individual factors to understand how these interact with societal factors to lead to perpetration. The highly gendered nature of sexual violence, however, in that most perpetrators are men and victims are more likely to be women, means analysis of gender structures as a basis for enabling sexual violence is necessary.

Jewkes (2012) noted that it is gendered inequities in social roles that appear to enable sexual and physical violence against women. This study was used for considering whether attitudes play an important role in the greater environment as enabling factors for perpetration in general, without them specifically working to differentiate men who have not perpetrated sexual violence (Jewkes, 2012). Included in this are ideas around men's authority over women and beliefs about sexual entitlement that saw men's motivations to rape include entitlement, "fun" (i.e., rape as entertainment) and punishment. In a multi-country study of factors associated with non-partner rape perpetration, a key finding was that a sense of sexual entitlement was a frequently given reason from men who had raped a non-partner woman (Jewkes, 2012).

The idea that rape is associated with the exertion of power is not new, but quantitative findings that link rape perpetration to certain behaviours, such as having a higher number of sexual partners, having sex with sex workers and using physical violence against a partner, appear to tie-in with a conceptualisation of masculinity that emphasises heterosexual performance and the use of that power over women (Fulu et al., 2013).

An ongoing difficulty with research in the areas of sexual violence and intimate partner violence is the limitations on comparability between settings because of variations such as data availability or definition, as well as rates of disclosure, which can be affected by various factors including cultural differences and interviewing techniques.

The importance of attitudes about violence against women as an indicator of gender inequality

Attitudes about the legitimacy of violence against women are an important indicator of widespread societal acceptance or otherwise of violence against women (Fanslow et al., 2010). This is particularly relevant in a cross-cultural context because variations in attitudes can indicate a variation in acceptance of violence against women and could be considered against prevalence indicators to gain an insight into the connection between violence and equality. Attitudes that are held about violence can lead to other factors, however, such as a willingness to report violence in societies where violence is not socially acceptable. This could potentially impact on data that uses rates of reported violence.

A VicHealth study that researched factors that influence community attitudes about violence against women included findings that "attitudes to violence against women are inextricably grounded in and intertwined with attitudes towards women, gender and sexuality" (Victorian Health Promotion Foundation et al., 2006, p. 6).

In addition, community judgements about violence against women are influenced by more general attitudes to gender and sexuality (Victorian Health Promotion Foundation et al., 2006). These attitudes vary within different cultural contexts and suggest that socio-cultural factors may be more important than gender alone in influencing attitudes to violence against women These attitudes reflect gender norms and social beliefs about male and female roles and positions in society (Nayak et al., 2003).

As an indicator of equality, attitudes have to form one aspect of measurement because in a society where women are considered of equal status to men, this would have to impact on the societal level of the ecological model in terms of beliefs that support violence against women.

Other perspectives

These considerations of culture and other factors are incorporated into perspectives that challenge gender inequality as the principle underlying factor of violence against women. Sokoloff and Dupont (2005) noted that there are various groups of advocates challenging the traditional feminist view that gender inequality is the key factor underlying family violence. They argue that gendered power is modified by other compounding intersections of power such as race and class. Same-sex intimate partner abuse is sometimes held up as an example whereby gender as an explanation alone is insufficient, however there may be additional power and control pressures on same-sex victims - fear of homophobic responses to reporting the abuse and fear of discrimination or invisibility - leading them to stay in unsatisfactory relationships (Sokoloff & Dupont, 2005).

Another perspective that raises questions about gender as the prime determinant in a family violence sense is that victim survivors may experience additional power deficits in terms of homelessness, economic challenge and racism (Sokoloff & Dupont, 2005). This is one of the key arguments against in debates about gender mainstreaming as policy.

Issues in the implementation of gender equality policies

The discussion in this section relates mainly to the strategy of gender mainstreaming that is now a key plank of institutional promotion of gender equality at the organisational level. It involves the incorporation of gender issues into the mainstream management of an organisation and requires analysis of gender in respect of all facets of that body's processes and policies (WHO, 2009). In a theoretical sense, gender mainstreaming is the process of reconsidering concepts and definitions to acknowledge that the world operates in a way that is impacted by differences in genders and to conceptualise and allow for these differences (Walby, 2005).

Feminist concerns regarding the process of gender mainstreaming include apprehension about what success will look like in terms of equality, (i.e. is it sameness or accommodation of difference?) The problem is that separate gender norms and standards mean that differences could be absorbed into an equality that is set at the standards valued by the status quo - male favoured structures. This has been identified as a concern in that the mainstreaming of gender issues detracts from the primacy of issues that females face that are different to men. Mainstreaming could result in men and women being treated the same, with the indirect effect of disadvantaging women (Pease, 2008). In Australia, this was seen in a policy context whereby women-specific services and policy were closed as unnecessary (Pease, 2008); however, such actions clearly miss the point of gender mainstreaming.

An alternative approach is to consider whether female gender norms and standards can instead be equally valued (Flood, Fergus, & Heenan, 2009). Walby (2005) noted that even if the goal of gender mainstreaming is gender equality, there is disagreement as to how this is done and what it would look like - whether it means accepting and valuing existing gendered differences? Her analysis noted that gendered domains that have the capacity to impact on gender equality, for example employment, family care, education, need to be examined to understand whether changes in one will have ripple effects and implications for the other domains.

Gender and intersectionality with other types of disadvantage

One of the key criticisms of gender mainstreaming is that it is limiting in its ability to incorporate other forms of inequality by which people are grouped socially such as race and economic disadvantage. Gender becomes the main perspective of application of equality without acknowledging that this can be a homogenising process that doesn't appreciate the complexity of other types of compounding disadvantage. Intersectionality describes theoretical analysis of power imbalances occurring across multiple systems of advantage and disadvantage. In this way, no single dimension of social inequality or social characterisation, for example gender, is prioritised.

The technicalities of implementing gender mainstreaming are also a subject of debate. It is easier to talk about incorporating gender mainstreaming while not considering the expertise that is likely to be required to implement it. Where policy-makers and political players have a high level of understanding, the implementation of gender mainstreaming strategies will be less problematic (Walby, 2005). But how will strategies be implemented in an environment where this is not the case and where the input of expertise from sources external to government may be required?

Measuring gender equality

What is success in terms of reaching gender equality? How can a society be considered gender equal? In the gender/violence against women literature, it is broadly theorised that increased gender equality will lead to decreased violence against women (VicHealth, 2007; WHO, 2010; Pease, 2008). This is also known as the ameliorative theory and equates to a sense that greater gender equality will ameliorate the amount of violence against women (Martin, Vieraitis, & Britto, 2006; Whaley, Messner, & Veysey, 2011). Gender equality can be an imprecise concept. Generally, it could mean that the work of men and women is more equally valued, family responsibilities more shared, political and social power less divided and men having less power over women (Whaley et al., 2011). But it will have different meanings depending on the different perspectives of those considering gender equality. For example, the international Organisation for Economic Co-operation and Development (OECD) looks at gender equality through a prism of economic growth and thus concentrates its measures on education, employment and entrepreneurship of each gender. However, what we are investigating are actually social causes of inequality and it is important that sociological perspectives are incorporated into understanding the causes of difference in these economic measures.

The importance of social status

What to measure and how to measure it are two very clear concerns for understanding gender equality changes. For example, one challenge in measuring different aspects of gender equality is being able to analyse the interaction between the economy in terms of paid workers and the "hidden" contribution that unpaid, family care makes to this. Also potentially affecting measurement of gender equality is the possibility of "backlash", which refers to an increase in male violence against women as women's status increases and men seek to maintain their hold on power.

Gender equality and empowerment of women was goal 3 of the United Nations Millennium Development Goals and therefore has been the subject of considerable dialogue regarding measurement. The official indicators for the UN measurement of this goal are:

• the ratio of girls to boys in primary secondary and tertiary education;
• the share of women in waged employment in the non-agricultural sector; and
• the proportion of seats held by women in national parliament.

These are necessarily very broad goals and can't really demonstrate the progress towards equality at other levels of the ecological social system. Individual cultures and resources will obviously have an impact on the rate of progress. The UN collates the Gender Inequality Index that rates countries according to the measures of inequality mentioned above and additionally, maternal mortality rates and adolescent fertility rates.

In Australia, there are a range of statistical measures developed by the Australian Bureau of Statistics. These are referred to as gender indicators and the measures include: economic security, education, health, work and family balance, safety and justice, and democracy, governance and citizenship.

In a paper considering methodologies for measuring gender equality, Scott (2010) observed that in order to understand gender equality thoroughly, it is necessary to understand how these inequalities are produced and reproduced over the life course and at different ecological levels. This means observation and measurement at different times periods, different levels and different cultures (Scott, 2010). The methodology used to measure and record these observations will affect the information that is produced and therefore what is understood about gender equality. This means, gender equality measures must be uniform to enable comparison, and the questions being asked about gender equality must be clearly framed in order to achieve the goals of the research (Scott, 2010). In addition, good quality indicators of gender equality that will yield the necessary information must be identified.

Women's status has been used as a measure of equality in the context of social constructions of value applied to each gender according to their roles and social arrangements within a culture (Bradley & Khor, 1993, Yodanis, 2004). Areas of measurement that are regularly understood to represent measures of equality and women's status include the spheres of education, occupation and political participation as per the United Nations Millennium Development Goal measures. In a persuasive argument for their particular conceptualisation of status as a measure, Bradley and Khor (1993) argued that status must not only be considered in terms of the dimensions of activity such as economic, political and social but rather in relation to the private and public domains which affect women's lives. The private dimension includes such measures as violence in the home, control over reproductive power, kinship and care, and social obligations. The public dimension refers to institutions and activities that are the measures most commonly used in reference to gender equality (Bradley & Khor, 1993). This is a similar conclusion to the aforementioned work of Levinson (1989) that indicated different spheres have different impacts on violence.

In terms of linking the measures of gender equality to violence against women, Yodonis (2004) pointed out that it has been very difficult to comprehensively test the theory of gender equality and gendered violence. In her study looking at societal level measures of gender inequality and rates of violence against women and women's fear relative to men's, she used data from the UN and an International Crime Victims Survey in the US. Although not a standard measure, when looking at gender equality, the fear women feel, relative to men's fear, whether justified or not, is a mechanism that can be used to indicate perceived vulnerability to men. She measured the status of women and linked it to violence by measuring the relationship between structural gender inequality, experiences of sexual and physical violence against women and levels of fear (relative to men's) among women in European and North American countries. Dimensions of measurement used were education, occupational and political status. Her findings supported the theory that the status of women was related to the prevalence of violence against women, and more strongly with sexual violence than physical violence. The overall prevalence of sexual violence was also linked to a higher rate of women's fear relative to men's. Yodonis also proposed the possibility of using women in legal and criminal justice systems as further measures of the status of women.

As noted earlier, Ridgeway (2014) makes the observation that despite changes in structural inequality in society, which have improved women's socio-economic and public sphere status or absolute status, their position as inferior in terms of social status has remained constant. This is an important observation. Her theory is that social status beliefs are a motivator of human behaviour. Group connections, information exchange, affection and beliefs about competence are perpetuated throughout social encounters and organisations that distribute power and resources, intrinsically directing higher status groups towards privileges. This type of thinking emphasises the importance of community attitude information being incorporated into measures of equality as much as socio-economic, political and other "hard" measurement data. Ridgeway and Correll (2004) concluded that although changing women's economic disadvantage and making changes at the personal and community levels do modify cultural beliefs about gender, the core hierarchy of these beliefs is more difficult to break down and needs to be persistently challenged at all levels.

Conclusion

Within the research and literature around violence against women it seems that all roads eventually lead to the need to readjust structural power imbalances to alleviate the greater privileges men have. However, the lack of clear and definite conclusions about which aspects of this power and how it drives perpetrators of violence indicates that more research is needed into the broad, societal level factors that underlie violence against women. Observations around the persistence of gender hierarchies that accord men greater status despite extensive social changes must be understood.

Consideration of the relative social status of the genders has implications for prevention of violence against women strategies and policies at all levels of the ecological model including educational programming and service delivery.

• The goals of prevention, reduction and elimination of intimate partner and sexual violence are currently being approached with a focus on achieving gender equality but more research is needed to understand exactly what this means and how it can best be implemented at the societal level. Incorporating sociological perspectives that enhance our understanding of which aspects of inequality have the most impact on violence against women will be beneficial for informing the policy around this important area of work.
• Measures and economic indicators around gender equality should include consideration of societal attitudes and structural hierarchies and how these affect access to resources and power.
• In education and programming to reduce violence against women, an understanding of gender and its relationship to status must inform education and programming to ensure esteem and respect for differences between socially constructed groupings. Constructs of masculinity and femininity, so defined by cultures and social norms, need work to accommodate a masculinity that does not include violence and that incorporates attitudes of mutual respect between the genders. The evidence base around how to do this is currently expanding.
• Organisational change could incorporate policies and work practices that acknowledge the impacts of gender difference and attitudes towards that difference in order to challenge traditional norms around gendered social roles.
• At the service provision level, recognising and acknowledging the impacts that entrenched inequality based on gender and other intersecting disadvantage have is important in order to move forward.

References

• Archer, J. (2006). Cross-cultural differences in physical aggression between partners: A social-role analysis. Personality and social psychology review, 10(2), 1331-1353.
• Bradley, K., & Khor, D. (1993). Toward an integration of theory and research on the status of women. Gender and Society, 7(3), 3473-3478.
• Fanslow, J., Robinson, E., Crengle, S., & Perese, L. (2010). Juxtaposing beliefs and reality: Prevalence rates of intimate partner violence and attitudes to violence and gender roles reported by New Zealand women. Violence against Women, 16(7), 8128-8131.
• Flood, M., Fergus, L., & Heenan, M. (2009). Respectful relationships education: Violence prevention and respectful relationships in Victorian Schools. Melbourne: Department of Education and Early Childhood Development.
• Fulu, F., Warner, X., Miedema, S., Jewkes, R., Roselli, T., & Lang, J. (2013). Why do some men use violence against women and how can we prevent it? Quantitative findings from the UN multi-country study on men and violence in Asia and the Pacific. Bangkok: UNDP, UNFPA, UN Women and UNV.
• Hankivsky, O., & Christoffersen, A. (2008). Intersectionality and the determinants of health: A Canadian perspective. Critical Public Health, 18(3), 2712-2783.
• Heise, L. (1998). Violence against women, an integrated, ecological framework. Violence Against Women, 4(4), 2622-2690.
• Jewkes, R. (2002). Intimate partner violence: causes and prevention. The Lancet, 359(April 20), 14231-14429.
• Jewkes, R. (2012). Rape perpetration: A review. Pretoria: Sexual Violence Research Initiative.
• Krug, E., Dahlberg, L., Mercy, J., Zwi, A., & Lozano, R. (2002). World report on violence and health. Geneva: World Health Organization.
• Levinson, D. (1989). Family violence in cross-cultural perspective. (Frontiers of Anthropology, vol. 1) Newbury Park, Calif: Sage Publications,
• Martin, K., Vieraitis, L., & Britto, S. (2006). Gender equality and women's absolute status: A test of the feminist models of rape. Violence against Women, 12(4), 3213-3239.
• Nayak, M., Byrne, C., Martin, M., & Abraham, A. (2003). Attitudes towards violence against women: A cross-nation study. Sex Roles, 49(7/8), 3333-3342.
• Pease, B. (2008). Engaging men in men's violence prevention: Exploring the tensions, dilemmas and possibilities. (Issues Paper 17). Sydney: Australian Domestic and Family Violence Clearinghouse.
• Quadara, A., Nagy, V., Higgins, D., & Siegal, N. (2014). Conceptualising the prevention of sexual abuse: Final report to the Department of Social Services. Melbourne: Australian Institute of Family Studies.
• Quadara, A., & Wall, L. (2012). What is effective primary prevention in sexual assault? Translating the evidence for action (ACSSA Wrap No. 11). Melbourne: Australian Institute of Family Studies.
• Ridgeway, C. (2014). Why Status Matters for Inequality. [Presidential Address]. American Sociological Review, 79(1), 11-16.
• Ridgeway, C., & Correll, S. (2004). Unpacking the gender system: A theoretical perspective on gender beliefs and social relations. Gender and Society, 18.
• Scott, J. (2010). Quantitative methods and gender inequalities. International Journal of Social Research Methodology, 13(3), 2232-2236.
• Sokoloff, N., & Dupont, I. (2005). Domestic violence at the intersections of race, class and gender: Challenges and contributions to understanding violence against marginalised women in diverse communities. Violence Against Women, 11(1), 386-864.
• UN Office of the Special Advisor to the Secretary General on Gender Issues and the Advancement of Women (OSAGI). (n.d.). Important concepts underlying gender mainstreaming (Fact sheet No. 2). Retrieved from <www.un.org/womenwatch/osagi/conceptsandefinitions.htm>
• UNFPA United Nations Population Fund. (2008). Promoting gender equality: Frequently asked questions about gender. New York: UNFPA. Retrieved from <www.unfpa.org/gender/resources_faq.htm>.
• VicHealth. (2007). Preventing violence before it occurs: A framework and background paper to guide the primary prevention of violence against women in Victoria. Melbourne: VicHealth.
• Victorian Health Promotion Foundation, Flood, M., & Pease, B. (2006). The factors influencing community attitudes in relation to violence against women: a critical review of the literature. Melbourne: Victorian Health Promotion Foundation.
• Walby, S. (2005). Gender mainstreaming: Productive tensions in theory and practice. Social Politics, 12(3), 3213-3243.
• Whaley, R., Messner, S., & Veysey, B. (2011). The relationship between gender equality and rates of inter and intra sexual lethal violence: An exploration of functional form. Justice Quarterly, 7327-7354.
• World Association for Sexual Health. (2008). Sexual health for the millennium: A declaration and technical document. Chapter II: Advance toward gender equality and equity. International Journal of Sexual Health, 20(Supplement 1), 293-295.
• World Health Organization. (2005). WHO multi-country study on women's health and domestic violence against women: Summary report of initial results on prevalence, health outcomes and women's responses. Geneva: WHO.
• World Health Organization. (2009). Strategy for integrating gender analysis and actions into the work of WHO Department of Gender, Women and Health. Geneva: WHO.
• World Health Organization. (2010). Preventing intimate partner and sexual violence against women: Taking action and generating evidence. Geneva: WHO/London School of Hygiene and Tropical Medicine.
• Yodanis, C. (2004). Gender inequality, violence against women and fear: A cross-national test of the feminist theory of violence against women. Journal of interpersonal violence, 19(6), 6556-6575.