Women's experiences of male violence : findings from the Australian component of the International Violence Against Women Survey (IVAWS)

Jenny Mouzos and Toni Makkai
ISBN 0 642 53842 5 ; ISSN 1326-6004
Canberra: Australian Institute of Criminology, 2004
(Research and public policy series, no. 56)

Chapter 1 : Introduction

It is generally acknowledged that the true extent of violence against women in Australia and elsewhere is unknown. While there are estimates, these are widely considered to underestimate the extent of victimisation. The factors that affect the estimates are as follows:

• most violence experienced by women occurs within the privacy of their own home and is less likely to come to the attention of others (Fineman & Mykitiuk 1994);
• even when the violence reaches the public realm, it is often viewed as less serious than violence that takes place outside of the home, between acquaintances and strangers (Cardarelli 1997);
• socio-cultural differences in terms of what is actually perceived and labelled as 'violence' mean that violence against women is not always acknowledged as a social problem, and more importantly, a problem that requires attention; and
• both administrative data collection systems and surveys are subject to a variety of biases, including a lack of comparability, that affect data on violence against women.

Despite the various limitations with developing national estimates two key findings have emerged from a range of research studies in this field. The first is that violence against women is 'more than an occasional, isolated incident ...' (Hanmer & Saunders 1993: 3). The second is that, the risk of victimisation is not the same for all women. The level of victimisation experienced by women varies based on their age, marital status, and most importantly, on the relationship, if any, that exists between themselves and the perpetrator of the violence (Mouzos 2003; Johnson 1996; Coumarelos & Allen 1998).

The most likely source of data on victimisation is from administrative police data files. However, violence against women is often not reported to police, leading others to conclude that such data represent only the 'tip of the iceberg' (Lievore 2003). Administrative collections rarely collect the level of detail surrounding the 'incident(s)' to enable more detailed analyses that could assist in effective crime prevention policies and initiatives (see Koss and Dinero 1989; Siegel & Williams 2003: 903). Lack of reliable data on the experiences of violence by women has hampered the formation of evidence led policy in this area.

As a result, in the last few decades, attention has shifted to understanding and documenting violence against women. The National Committee on Violence Against Women was convened in 1990 comprising representatives from Commonwealth and State and Territory Governments in Australia. This committee noted, 'it is impossible to provide a complete national profile of the number of women who are victims of violence' (1993: 1). In their report, they put forth a number of directions for action, including that Commonwealth agencies work together in developing an appropriate tool, which covers a range of indicators with respect to the levels of violence against women.

In 1997, the Australian Government introduced the Partnerships Against Domestic Violence (PADV) initiative which is a national framework for responding to both short and long term effects of domestic violence in the Australian community. Partnerships implements a collaborative approach between jurisdictions at a practical and policy level. Additional funding for continuing Partnerships was provided in 1999, focusing on community awareness; prevention and early intervention with children who witness or are the victims of domestic violence; and grass roots projects to strengthen Indigenous community efforts to address family violence.¹

Based on the experiences of countries such as the United States (Tjaden & Thoennes 1998; 2000; Straus & Gelles 1986), Canada (Statistics Canada 1993; Rodgers 1994; Johnson 1996) and New Zealand (Morris 1997), which had conducted nationally representative surveys, the first Australian survey of violence against women was conducted in 1996 (ABS 1996). While the results from this survey have been significant in increasing our knowledge base on violence against women, comparisons cannot be made between Australia and other countries.² To assess whether social reform policies are having an impact on the degree of violence experienced by women it is necessary to have ongoing monitoring over time to examine whether the level of physical and sexual victimisation of women has changed.

The Australian Government has invested $23.2 million through the establishment of the National Initiative to Combat Sexual Assault to reduce and prevent sexual assault. One of the objectives of the Initiative is to develop an information strategy that:

• facilitates access to national policy relevant data on sexual assault;
• informs effective responses in the prevention of sexual assault; and
• establishes a comprehensive evidence base incorporating both research and practice.

To this end, the Australian Government's Office of the Status of Women funded Australia's participation in the first International Violence Against Women Survey (IVAWS), as well as another national survey on violence - the ABS Personal Safety Survey - due to run in 2005/2006. These initiatives will complement existing data sources on violence against women and assist in the creation of evidence led policies. This report therefore seeks to inform the strategy by providing the most up to date information on physical and sexual violence experienced by over 6000 women in Australia.

Definitional issues: measuring violence against women

As many researchers in the field will attest, measuring violence is not straightforward (Lobmann et al. 2003; Stanko 1990). Considerable debate occurs because researchers have failed to agree on such basic matters as a definition:

Researchers have used terms related to violence against women in different ways and have used different terms to describe the same acts. Not surprisingly, these inconsistencies have contributed to varied conclusions about the incidence and prevalence of violence against women (Saltzman, Fanslow, McMahon & Shelley 1999: 1).

Ambiguity exists in terms of 'what do we mean by violence against women'? This question can be divided into two parts: (1) what do we mean by 'violence'; and (2) what do we mean by 'violence against women'.

What do we mean by 'violence'?

In the narrow sense, violence is generally understood as an incident in which an individual intentionally injures another person. However, such a narrow definition does not include the 'non-contact' types of violence, such as psychological and emotional violence or even threats to use violence. In order to obtain a complete picture of the manifestations of violence experienced by women in Australia (and elsewhere), it is important that both actual violence and threats of violence are included in the measurement of violence.

The IVAWS measures three distinctive types of violence against women (see the glossary of terms at the beginning of this report):

1. physical (including threats of physical violence);
2. sexual (including unwanted sexual touching); and
3. psychological (controlling behaviours).

Physical violence includes both physical assaults, which refer to the use of physical force with the intent to harm or frighten a woman, and physical attempts or threats, which refers to the verbal, and/or physical intent to inflict harm which the woman believed was able and likely to be carried out. The IVAWS measures the following types of physical violence:

• having something thrown at her or hit with something;
• being pushed or grabbed or having her arm twisted or hair pulled;
• being slapped, kicked, bit or hit with a fist;
• attempt to strangle or suffocate or burn or scald on purpose;
• having a gun or knife used or attempted to be used on her; and
• any other physical violence.

Sexual violence is any form of non-consensual or forced sexual activity or touching including rape. It is carried out against the woman's consent using physical or threatened force, intimidation or coercion. The IVAWS measures the following types of sexual violence:

• being forced into sexual intercourse;
• attempted to be forced into sexual intercourse;
• being touched sexually in a way that was distressing;
• being sexually assaulted whilst under the influence of drugs³ that she was given without her knowledge so that she could not give her consent;
• being forced or attempted to be forced into sexual activity with someone else, including being forced to have sex for money or in exchange for goods; and
• any other sexual violence.

Psychological (or emotional) violence includes insults, humiliation, put-downs, restrictions of freedom and constant surveillance. Many of these are known also as controlling behaviours. One of the most common examples of psychological violence against women is intimate partner psychological abuse. The IVAWS measures the following controlling behaviours of male intimate partners:

• restricted or tried to prevent contact with other men by becoming angry if she spoke with other men;
• was not supportive if she engaged in activities outside the home;
• tried to limit her contact with family or friends;
• follows her or keeps track of her whereabouts;
• calls her names, insults her or behaves in a way to put her down or to make her feel bad; and
• damages or destroys her possessions or property.

What do we mean by 'violence against women'?

The United Nations Declaration on the Elimination of Violence Against Women General Assembly Resolution 48/104 (20 December 1993: 2) defines violence against women as:

... any act of gender-based violence that results in, or is likely to result in, physical, sexual or psychological harm or suffering to women including threats of such acts, coercion or arbitrary deprivations of liberty, whether occurring in public or private life.

It has been suggested that the term 'violence against women' is even more difficult to define (Lobmann et al. 2003). Ambiguity centres on the question of whether being a female victim is sufficient to classify the aggressive act as 'violence against women'? Consider the following example:

Let us suppose that a burglar does not know if the occupier of a home is male or female. Then a residential break-in would surely not be a crime against women, even if the victim were female (Lobmann 2003: 310).

In the example provided above, the victim was not specifically targeted because she was a female, in other words, because of her biological sex. There are some offences where the victim's biological sex is fundamental to the offence, for example, rape. A heterosexual rapist would seek victims of the opposite sex. 'In this case, the victim's sex is a (necessary) condition for the offence to be committed' (Lobmann et al. 2003: 311). If however the rapist was targeting women as a means of asserting his male superiority and power over women, then it is not the women's biological sex but gender or social role that makes the act specifically violence directed against women.

Therefore, the concept of 'violence against women' as we understand it is the intentional physical, sexual and/or psychological abuse of women due to their biological sex and/or their social role (Lobmann et al. 2003: 311).

What is IVAWS?

The International Violence Against Women Survey (IVAWS) is an international comparative survey designed to collect information on women aged between 18 and 69 and their experiences with violence by males, and specifically to gauge the level of victimisation experienced by women in Australia and abroad. The survey is the result of collaborative efforts involving two United Nations criminal justice agencies - the United Nations Interregional Crime and Justice Research Institute, and the European Institute for Crime Prevention and Control - and Statistics Canada. Australia is the first country to fully implement the project after two extensive pilots.

The methodology for the IVAWS evolved out of the International Crime Victim Survey (ICVS); a survey with a broad focus on crime victimisation (Carcach & Makkai 2003). ICVS was not designed to target violence against women. An international survey that would facilitate comparative analyses on violence against women was therefore required. Utilising the comparative methodology of the ICVS, as well as other national violence against women surveys, the IVAWS was developed to focus specifically on women's experiences of violence by males and to allow for international comparisons.

The aim of IVAWS is to:

• promote research on violence against women in developed and developing countries;
• create an international data set for cross-cultural analysis of violence against women around the world; and
• promote policy development for preventing and counteracting violence against women both nationally and internationally.

Objectives of IVAWS

The main objectives of IVAWS are to:

• contribute information for the development of legislative measures and means of victim assistance;
• promote community and victim-centred prevention strategies;
• provide information to assist in the development of evidence led policies;
• facilitate international research;
• improve police practices in dealing with cases of violence against women, by:
  1. revealing the extent of unreported crimes; and
  2. exploring the level of satisfaction with the police and their response; and
• formulate and test research hypotheses and theories.

The survey collects data about women's experiences of the following types of violence by males:

• non-partner violence;
• current intimate partner violence;
• previous intimate partner violence; and
• parental abuse history and childhood victimisation.

The survey also collects detailed information concerning:

• the impact of violence on women, such as, injury and fear for life;
• women's perceptions of male violence, such as, seriousness of the incident, and whether it is considered to be a crime; and
• women's reactions to violence, such as, contacting specialised agencies or reporting to police.

Focus of this report

The focus of this report is on providing estimates of the nature and extent of violence experienced by a sample of Australian women. It explores the different forms of violence experienced by these women, and provides a profile of the males who committed the violence. This exploration is based on the analyses of responses from 6,677 women aged between 18 and 69 from across Australia. These women were randomly selected and interviewed by telephone for the IVAWS during 2002-2003. It is important to acknowledge that the data are drawn from a sample and any estimates and probabilities, will have sampling error. However, the relatively large sample size works to minimise the extent of the error.

This report is divided into six main chapters. The following chapter provides an overall account of women's experiences of violence in Australia. This includes the occurrence of violence (physical and sexual) experienced by women over their lifetime (that is, since the age of 16), and in the five years and 12 months preceding the survey. It also examines the characteristics of those women who experienced some form of violence (for example, their age group, marital status, educational level, minority population background status (Indigenous status, non-English speaking background) and employment status), and the severity of violence, including the type of injury sustained.

The third chapter focuses specifically on women's experiences of intimate partner violence (current and previous marital, de facto partners and boyfriends), occurring throughout their lifetime as well as in the past 5 years and 12 months. The characteristics of the most recent incident of intimate partner violence are explored, including the severity of violence experienced, details relating to the incident, such as where it occurred, and drug or alcohol involvement. Included in this section are the correlates of victimisation and the characteristics of the male partners.

The fourth chapter examines women's experiences of non-partner violence (such as violence from family members, friends, acquaintances and strangers) across the adult lifetime and in the past five years and 12 months. The characteristics of the most recent incident of non-partner violence are explored, including where the incident occurred, how many men were involved in the incident, drug or alcohol involvement and the type of injury that women sustained. As with the previous section, the correlates of victimisation and the characteristics of the non-partners are analysed.

The next chapter focuses on women who reported being physically or sexually abused before the age of 16 years, by whom they were victimised, and whether childhood victimisation is a significant risk factor for adult victimisation. In other words, is there any evidence of a generational cycle of violence?

The sixth chapter focuses specifically on women who have experienced violence and their perceptions as to the seriousness of the violence, and whether they considered it a crime. It also examines women's reactions to violence, such as whether they reported to police or other specialised agencies, and their levels of satisfaction with how the matter was dealt with. Reasons for non-reporting are also presented.

Methodological issues

In conducting surveys, such as the IVAWS, comparability between surveys is affected by a number of factors. In Australia, there are potentially two other key sources of national data on violence against women:

1. The Crime and Safety Survey; and
2. The Women's Safety Survey.

These surveys differ from IVAWS in a number of ways (see Appendix 2 for a brief comparison of these surveys):

• the time-periods examined differ;
• IVAWS captures information on a greater variety of physical and sexual violence that women may experience, consequently this results in higher overall estimates of violence;
• the wording of questions, definitions, and the sequence of questions differ;
• administration of the questionnaire (telephone versus face-to-face interviewing) differs; and
• the surveys frame and introduce screening questions for each of the sections in different ways, which can significantly affect the level of disclosure (see Tjaden & Thoennes 2000: 31).

Given these differences, it is strongly advised against drawing definitive inferences from IVAWS with the results from previous surveys.

As mentioned above, the IVAWS was a telephone survey and participation in the survey was therefore limited to women living in households with telephones. Consequently, the survey does not reflect the experiences of women living in households without telephones, women who are homeless or living in group facilities or institutions. In addition, with the growing use of mobile telephones4, more people, especially those in the younger age brackets may be opting for the use of mobile telephones only as opposed to a landline telephone. This may result in an under representation of women who do not have a landline telephone, as they were omitted from the survey selection process (see Appendix 1).

The IVAWS sample has been weighted to represent the population of women aged between 18 and 69 by age and geographic area. As mentioned the estimates are subject to sampling variability because they are derived from a sample of the population and not the whole population. In other words, the estimates may differ from results that would have been produced if all women aged 18 to 69 in Australia had been surveyed. Where appropriate, confidence intervals, the standard error and the relative standard error (RSE) of the estimates have been calculated. The RSE of the estimate indicates the extent to which an estimate might have varied because a sample was used (the RSE is the standard error divided by the actual point estimate expressed as a percentage).

The standard error of the mean (or estimate) indicates how close the estimate is to the true population mean (plus or minus two standard errors produces the range of the sample mean of the population). For example, two standard errors of an estimate of 48 per cent based on a sample of 6,677 is about one per cent meaning that there is a 19 in 20 chance the true population value lies between 47 per cent and 49 per cent. This range is referred to as a confidence interval.

Consistent with national standards, estimates with relative standard errors of over 25 per cent are considered unreliable (ABS 1996: 74). While estimates with higher relative standard errors are presented in this report, caution should be exercised when referring to these estimates. Throughout the report, a single asterisk (*) denotes estimates with a relative standard error greater than 25 per cent. In these cases estimates are not provided. In addition, as estimates presented in this report generally exclude 'don't know/can't remember', 'refused/no answer' and other invalid responses, sample and sub-sample sizes (n) will vary throughout the report.

Due to rounding some percentages may not always sum to zero. All confidence intervals (CI) are at the 95 per cent level.

Finally, as a voluntary survey women could and did refuse to participate. Appendix 1 details the survey methodology. In summary, 17,247 households were contacted by telephone and 6,677 women agreed to participate resulting in an effective participation rate of 39 per cent. Although this sample of women has been weighted to reflect the overall age and geographic location of all women aged between 18 and 69 in Australia it is possible that some unknown bias could be affecting the results.

---

• ¹ For additional information on PADV see: http://ofw.facs.gov.au/womens_safety_agenda/previous_initiatives/padv/index.htm
• ² The surveys in the United States and Canada were conducted in different time periods and used different definitions of violence.
• ³ This refers to both illicit and licit (prescription) drugs.
• ⁴ The number of mobile phones in Australia outnumbered fixed lines in 2002 (Herald Sun, "Mobiles could corner the market", 03/05/04).