New Zealand National Survey of Crime Victims 1996
Women's Safety Survey 1996
Sector and Agency Responses

Introduction

The 1996 New Zealand National Survey of Crime Victims is the first stand-alone national general victimisation survey to be conducted in this country.

The major purpose of victimisation surveys is to establish whether respondents or their households have been the victims of any of a range of offences over a specified period of time. Surveys of this type may also include supplementary surveys which focus on particular types of victims, and in this victimisation survey a sub-set of 500 of the female respondents were reinterviewed about male partner violence for the Women's Safety Survey.

The survey was commissioned and funded by the New Zealand Police, the Ministry of Justice, the Crime Prevention Unit of the Department of the Prime Minister and Cabinet, the Department of Social Welfare, Te Puni Kokiri, the Ministry of Women's Affairs and the Ministry of Youth Affairs. It was undertaken by Victoria Link Ltd (the research company of Victoria University) in conjunction with AGB McNair Ltd and Massey University's Statistics Research and Consulting Centre.

The survey was commissioned because there was a clear need for improved measures of the incidence and prevalence of victimisation in New Zealand. There was also a need for more information about the effects of, and responses to, victimisation. The survey provides valuable base-line data on which it is hoped to build by conducting subsequent surveys.

The results from the National Survey of Crime Victims and the Women's Safety Survey will inform planning and policy making in relation to government strategies and programmes. Policy development areas where the data from the survey will prove useful include responses to crime, family violence, responsiveness to Maori, services to victims, the criminal justice system, young people, and responsiveness to Pacific Island people.

Crime victimisation surveys, self-report studies and official police and court statistics are the three main ways of obtaining information on crime and crime rates. Each of these, however, is an incomplete measure of crime and crime rates owing to methodological problems with respect to, and limitations in the scope of, the data collected. With crime victimisation surveys, there may be concerns about the accuracy and validity of respondents' recall. Respondents may forget incidents, may refuse to tell interviewers about crimes (especially sensitive or traumatic crimes), may incorrectly define incidents as crimes, or they may report crimes which are outside the specified time period for the survey. As with any sample survey of the community, there can also be problems with sampling and sample error, depending on the sampling method and the sample size. These factors need to be taken into account when considering the survey findings.

Nonetheless, the findings from the crime victimisation surveys contribute to a more complete picture of crime and victims than would otherwise be possible.

This introduction is followed by the executive summaries from the final reports on the National Survey of Crime Victims and the Women's Safety Survey together with initial government sector responses to the findings of the research. The full reports of the National Survey of Crime Victims and the Women's Safety Survey, which include detailed analyses and discussion on the survey findings, are available from Bennetts Bookshops, nationwide, or by using the order form on page 11 of this summary.

New Zealand National Survey of Crime Victims 1996

This report presents the findings from the first comprehensive national survey of crime victims, comprising interviews with a random sample of the population aged 15 and over. The survey is designed to provide an alternative measure to police statistics of crime victimisation; to identify the extent to which the risks of victimisation vary between social groups; to provide information about the circumstances and impact of offences; to describe the effects of crime and victims' response to it; and to provide a range of other crime-related information.

The Nature and Extent of Crime

In general, the survey found that there were an estimated 2 million offences (including attempts but excluding commercial and business offences) against households and against individuals aged 15 and over in New Zealand during the 1995 calendar year and that only a small proportion of the offences disclosed in the survey (less than 13%) were recorded by the Police.

Violent offending and sexual offending, including threats, made up almost two-thirds of the total offences disclosed in the survey. However, many of the violent offences involved threats of violence or threats of damage to property, which clearly vary enormously in seriousness and significance; and many were sexual offences which had a large sampling error. If we exclude threats and sexual offences from our count, assaults, wounding and robbery make up about a quarter of the total offences, only slightly above the figure reported in the 1996 British Crime Survey.

For grievous assaults, other assaults and threats, we distinguished between violence by those known well to the victim (including family violence) and violence inflicted by casual acquaintances or strangers. Grievous assaults were roughly evenly divided between the two groups. Other assaults and threats, on the other hand, were almost three times more likely to be committed by those known well to the victim than by strangers or casual acquaintances.

There are essentially two reasons for the gap between police statistics and the number of offences disclosed in the survey: not all offences come to the notice of the police; and of those that do, not all are recorded by them.

In this survey only a little over 40% of the cases in which we were able to collect information on reporting came to the notice of the police. There was considerable variability in reporting rates between one offence and another, with nearly 90% of theft or unlawful taking of motor vehicles being reported, and only a quarter of damage offences and a third of assault offences. The willingness of victims to report crime was primarily dictated by the seriousness of the offence in question - its intrusiveness, the degree of threat involved, the extent of injury, damage or loss and, to some extent, its emotional consequences for the victim. However, other factors - in particular, the relationship between the victim and the offender - also influenced the reporting decision. It is also clear that some social groups were less willing to report offences than others: males and young people were less likely to report all offences, and Pacific Island respondents were less likely to report violence.

Of those that did come to the notice of the police, a considerable number were not recorded by them. Again, the pattern varied between offences. Approximately 80% of reported burglaries appeared to get recorded as such, by comparison with less than half of the reported damage offences and less than one third of reported thefts from inside or outside the home. There are a number of possible reasons for the fact that some reported offences were not recorded: the police may have regarded the complaint as fabricated or mistaken; they may have thought the evidence insufficient to substantiate an offence; they may have regarded it as too trivial to warrant the paperwork; or the incident may have been coded as one type of offence in our survey but recorded by the police as a different offence.

Although the survey findings indicate that there is much more offending than is disclosed by the official statistics, it would be wrong to jump to the conclusion that it is desirable that a greater proportion of offending be reported to the police. It is true that a considerable number of apparently serious offences - including wounding and sexual violation - went unreported to or unrecorded by the police. It is also noteworthy that some groups appeared more reluctant to report than others. In at least some of these cases, reporting may have resulted in the detection of a serious or persistent offender, or at least the provision of much needed support for the victim. However, the offending which needs to come to the notice of the criminal justice system is in fact only a small proportion of the range of behaviour which could potentially do so, and the fact is that the vast majority of the offences which went unreported or unrecorded were perceived by victims to have been relatively trivial and were thus handled in other ways. Victims' perceptions in this respect can, of course, change. The important implication of the survey data, therefore, is that overall trends in official crime statistics should be regarded with caution, since they are highly susceptible to changes in reporting and recording practices.

The Distribution of Victimisation

The total offences disclosed in the survey were not evenly distributed: some victims were repeatedly victimised. The distribution of violent offending was particularly heavily skewed. Thus only 0.5% of the total sample (or 6% of the victims of violent and sexual offences) were subject to five or more such offences, but they accounted for a massive 68% of such offending. Amongst such victims, the average number of violent and sexual offences was 12.

This finding has profound implications for crime prevention. If victimisation, especially violent victimisation, is concentrated in small pockets of the population, then (unless crime displacement occurs) substantial crime reductions are likely to be achieved by focusing crime prevention efforts on those who are particularly at risk of multiple victimisation. Even if some crime is displaced to other people who would otherwise have been non-victims, which is unlikely at least in relation to violent victimisation, the consequence of crime prevention efforts targeted at repeat victimisation would be at least to spread crime more evenly throughout the population and thus to dilute its impact on any one individual.

The risk of victimisation also differed between social groups. In relation to gender, about the same proportion of women as men were subject to some form of violent or sexual offending on one or more occasions (the prevalence rate). Moreover, while the average number of such offences to which women were subject (the incidence rate) was much greater, this gender difference did not reach statistical significance. The reason for this is that the higher incidence rate amongst women resulted almost solely from the fact that a small number of women were subject to a large number of repeated sexual attacks; because the sample of such multiple victims was small, the standard error (ie the sampling error) was very large. Nevertheless, it seems likely that, given the greater risk of sexual violence to which women are exposed, the pattern we have observed would be repeated and would reach statistical significance in a larger sample. In relation to other assaults, there were no significant gender differences, although it appeared that the nature of the assaults differed: the incidence and prevalence rates for assaults by strangers and casual acquaintances against men were much higher than those against women; conversely, women were much more likely than men to be assaulted by those they knew well.

There were differences in the overall victimisation rates of different age groups, with a general pattern of decreasing incidence and prevalence with increasing age. In fact, with the exception of arson/wilful damage, this pattern was observed for every offence category. However, the differences did not always reach statistical significance. The main statistical effects were observed in relation to violent and sexual offending, general theft and individual property offending in aggregate.

Maori had significantly higher rates of both the prevalence and incidence of assaults and threats than New Zealand European/European. A similar pattern was also evident in relation to both the prevalence and incidence of indecent assault and the prevalence of sexual violation, although given the small sample size the differences were not statistically significant.

The same trends did not emerge in respect of individual property offences. While the incidence of general theft was roughly the same amongst New Zealand European/European as Maori, it was lower amongst Pacific Island respondents, and, when individual property offences were taken as a whole, both the prevalence and the incidence rates were higher amongst New Zealand European/European, with the difference between New Zealand European/European and Pacific Island respondents being statistically significant.

There was a slight tendency for the prevalence (but not the incidence) of violence to be inversely related to socio-economic status, but the relationship between socio-economic status and both the prevalence and the incidence of individual property offences was in the opposite direction. Although these differences did not reach levels of statistical significance, one would expect that if the sample size were increased they would, since both of these relationships are broadly in line with other surveys.

The distribution of the prevalence of violence and burglary was examined in more detail by the use of a method of multi-variate analysis called logistic regression, which enabled possible interactions between different socio-demographic variables to be taken into account.

In relation to violence, the analysis generally confirmed the picture already presented: there was an inverse relationship between age and risk; there was a tendency, which did not quite reach statistical significance, for Maori and Pacific Island respondents to be more at risk; and there was no relationship between gender and risk, reinforcing the conclusion that women and men are equally at risk of violence, although they are exposed to different types of violence. In addition, the analysis showed that marital status influenced risk, with those in the divorced/separated category having the highest risk.

In relation to burglary, only the frequency of going out at night turned out to be significantly associated with the risk of burglary, with those who never went out at night, not surprisingly, being less at risk. There was also a marginal trend for the prevalence of burglary to increase with household size. Other factors - such as socio-economic status, the nature of the household, whether the house was owned or rented, and respondents' perceptions of their neighbourhood - did not significantly affect prevalence.

Lifetime Prevalence of Sexual Violence or Partner Abuse

We also obtained data on the extent to which respondents had been victims of sexual violence or partner abuse at any time during their lives. As expected, women reported a much higher prevalence of both sexual victimisation and partner abuse during their lives than men. For example, 11.6% of women reported that someone had had, or had attempted, sexual intercourse with them against their will, compared with only 1.5% of men; and 15.3% of women who had ever been in a "partnership" reported some form of partner abuse, compared with only 7.3% of men. This gender difference persisted across all ethnic groups, although it was particularly marked for Maori.

However, contrary to expectation, there were no age differences in the lifetime prevalence of some forms of sexual violence or partner abuse, and victims in the younger age groups were more likely to have experienced multiple forms of sexual violence. There were no clear ethnic differences in the lifetime prevalence of sexual violence, but Maori disclosed much higher rates of partner abuse.

Victims' Satisfaction with the Police

The survey asked victims who reported offences to the police whether or not they were satisfied with the police response. In the main, it found that victims were happy with the support they received from the police. Most thought that the police were seen as polite and pleasant in their dealings with victims. However, a significant minority of victims - in particular, victims of violent offences (ie non-domestic assaults and threats), younger victims, students, those engaged in home duties, beneficiaries, and Maori and Pacific Island victims - were dissatisfied with the police response.

The major reasons for victims' dissatisfaction fell into two distinct categories: they thought that police officers appeared uninterested and that they provided inadequate feedback on the progress of the case; and they were concerned that the police had failed to catch the offender or to recover their property, or did not seem to be doing enough to investigate the offence. This suggests that the major sources of dissatisfaction over which the police have control (ie apart from the failure to catch the offender or to recover the victims' property) have to do with the presentational style of police staff. Paying attention to what victims say, being seen to be taking them seriously, providing more feedback on what is being done or, if nothing much can be done, explaining why, would help remedy this situation. Indeed, explanations as to why offenders are unlikely to be caught or why property is unlikely to be recovered might also go a long way towards offsetting victims' dissatisfaction with the perceived `failure' of the police in this area.

The Circumstances and Impact of Violence

The survey findings challenge many of the images of violence portrayed in the media. Even excluding family violence and violence by people known well to the victim (for which we did not have sufficient data), assaults and threats were as likely to occur at home and at work as on the street. Much of the violence was fairly minor in terms of injuries and items stolen: in most robberies, there were no injuries and nothing was taken; and in most assaults, even when injuries were inflicted, they were relatively minor and only infrequently involved more than bruises and scratches. Moreover, much of the violence was described by victims as not having much impact on them, with more than half of the victims of assaults, threats and robbery saying that they were "just a little affected" by the incident, and only a little more than 10% saying that they were "very much affected".

In more than half of the assaults and threats, the offender(s) were believed by their victim to be under the influence of alcohol. At the same time, 35% of assault victims, 16% of robbery victims, and 13% of victims of threats said that they were under the influence of drugs or alcohol at the time of the offence. Overall, this is suggestive of a greater link between offending and alcohol or drug use than between victimisation and alcohol or drug use. However, this should be regarded as a tentative conclusion. The data were derived from victims' assessments: there are reasons for victims to under-report their own involvement in alcohol or drugs; their assessment of the offender's situation some time after the event may not have been accurate; and many victims did not know whether or not the offender was affected by drugs or alcohol.

Excluding family violence and violence by people known well to the victims, the typical victims of violent incidents were young men, and they were more likely than other groups to have violence perpetrated against them in a pub or club. In contrast, women aged 25 to 39 were more likely than other groups to have their violence perpetrated against them in a home-based setting.

The Circumstances and Impact of Burglary

Where the time of the burglary could be pinpointed, marginally more took place in the evening or at night than during the morning or the afternoon. On a daily basis, the risks of burglary on the weekend were only slightly higher than during the week. A significant proportion (20%) of victims could not specify the time and/or the day when the burglary took place, suggesting that there were an appreciable number of burglaries where the victim was away on holiday or for the weekend.

In almost a third (31%) of the burglaries, the victim was able to say definitely that someone was at home at the time of the burglary. In 40% of these cases, the people at home became aware of the burglary at the time. However, confrontations with burglars were extremely rare.

Something was stolen in almost three-quarters of burglaries. In 18% of the burglaries where property was stolen, all or part of the property was recovered. In just over half the burglaries, the property stolen or damaged was covered by insurance, and in two-thirds of these the victim had made or intended to make a claim. Nevertheless, even after insurance had been paid, 26% of the households victimised said that they were over $1000 out of pocket. Furthermore, these were only the immediate costs: such households were likely to have incurred increased insurance costs as a result of the offence and may well have incurred further home protection costs as well.

Only 5% of victims reported no personal reaction to the burglary. In those cases where some reaction was recorded, 17% described the incident as having a strong effect on the household, 28% saw it as having quite a lot of effect and 53% saw it as having just a little effect. In general, the reactions reported were muted. Most victims were simply angry and/or shocked; reactions such as fear and disturbed sleep were rare.

In many burglaries, there was no forced entry. It is not surprising to find, therefore, that 25% of the burglary victims had no special security measures in place at the time of the offence. Of those who did report special security measures, most mentioned deadlocks or double locks, security chains, safety window latches and/or security bolts. Only 1% of victims claimed to belong to a Neighbourhood Support Group, or to be part of a group of neighbours keeping an eye on each other's property, suggesting that they did not see this as a security measure or an effective crime prevention device. Overall, the burglary victims in this survey reported lower levels of security than burglary victims in the 1996 British Crime Survey.

Meeting the Needs of Victims

Part of the New Zealand Crime Prevention Strategy is to `address the concerns of victims and potential victims' (Crime Prevention Unit, 1994). This involves taking steps to minimise the fear of crime which can have negative effects on people's lives, to prevent repeat victimisation and to aid victims' recovery from the effects of offending. This survey investigated both respondents' awareness of victim support services and their perception of those services in meeting this last objective.

Almost two-fifths of respondents had no knowledge of community services, apart from the police, which were available for victims of crime and almost a third mentioned only one. There were also clear differences in knowledge between various socio-economic groups: groups with significantly lower levels of awareness were men, Pacific Island and Maori respondents, older respondents and those from lower socio-economic status groups.

Where victims do report offences to the police, it is now recognised to be "good practice" that they should be provided with information about other agencies which may be able to help them deal with the aftermath of the offence. Sometimes, as with Victim Support, their name may be provided to that agency for further follow-up. In this survey, however, only 12% of those who had reported the offence to the police recalled being subsequently contacted by the organisation. More generally, very few victims had contact with any type of victims' services, including Victim Support. Where contact was made, the advice or help provided was to a large extent appreciated.

There is some evidence that there is a need for better targeting of support services: some of those contacted by support agencies had little need for such services; conversely, others who were significantly affected by the crime - at least some of whom were probably in need of such services - were missed; and some victims specifically stated that they wanted additional support or help but did not receive it.

Fear and Concern about Crime

The survey asked respondents a range of questions about crime in their neighbourhood, which were largely aimed at exploring their perception of the importance of crime as a social issue. A number of significant findings emerged from these questions. First, only one third of the respondents thought that crime was a problem in their own neighbourhood. Secondly, amongst those who did think this, they were much more concerned about burglary than about any other offence, with over three-quarters mentioning it as a problem. This provides some support for the greater attention which has been paid to burglary in policing priorities in recent years. Thirdly, there was no clear relationship between either the incidence or prevalence of victimisation and people's perceptions of crime problems, indicating that people's concerns about crime are not derived primarily from their own experience. Fourthly, people's perceptions of local problems do not focus exclusively or even primarily upon crime; they were often just as concerned or even more concerned about other features of the neighbourhood - uncontrolled dogs, teenagers hanging around, rubbish and litter lying about - which indicated that the neighbourhood was not homogeneous, cohesive or properly controlled. Finally, despite the frequent media coverage of rising crime rates in the official statistics, more than half of the respondents thought that crime had remained the same in their own neighbourhood, and only 4% thought that there was a lot more of it.

Respondents were also asked a series of questions about their fear of crime. In particular, they were asked how often they went out at night; whether they walked alone in their neighbourhood after dark and felt safe in doing so; how worried they were about being a victim of specific offences; and how worried they were about other accidents and misfortunes.

The majority of respondents said that they went out after dark; only 7% never did so and of those only a small minority cited crime-related reasons for this. The majority of respondents also said that they walked alone in their neighbourhood after dark and felt safe in doing so; and even most of those who did not, said that they would feel generally safe if they did so (although there were significant gender differences in this respect).

When respondents were asked how much they worried about being the victim of 11 specific offences, those who said that they were "very worried" comprised over 30% of the sample in relation to two offences, and about 20% in relation to another seven. However, this finding may suggest no more than that 20% of respondents experience a transient feeling of significant anxiety in particular situations rather than a permanent emotion of anxiety which preoccupies their lives.

Levels of fear about crime were generally no higher than anxieties about other things. Respondents tended to be just as worried about the prospect of illness, a road accident or an accident in the home as they were about burglary, assaults by strangers or car theft. Fear of crime, then, is nothing out of the ordinary: it is simply part of a broader pattern of anxieties which arise out of the threats and uncertainties of everyday life. However, levels of fear differed significantly from one socio-demographic group to another. In particular, women, younger people, Maori and Pacific Island respondents and lower socio-economic status groups expressed greater worry about crime.

In general, the survey's findings suggest that people's fears about crime are influenced by the risk they confront and the impact which crime is likely to have upon them. Thus the experience of recent victimisation impacted upon fear of crime, although this was stronger in relation to property offending than violent offending; and there was also a positive relationship between fear and the risk of victimisation. Where groups expressing greater levels of worry did not have a higher risk of actual victimisation than others, their greater levels of fear were explicable in terms of their greater vulnerability and other social disadvantage.

Crime Prevention Strategies

The survey gathered information on the strategies adopted by people to prevent household crime and personal victimisation and, to a lesser extent, on the effectiveness of those strategies. In relation to personal victimisation, it found that a large proportion of respondents took some precautionary measures to protect themselves against crime when they were out at night; that one in four sometimes carried a weapon (or something that could be used as a weapon) for the specific purpose of protecting themselves against crime when they were out at night; and that those reporting the highest levels of fear of personal victimisation - in particular, women and those aged 15-24 - were also more likely to take precautionary measures.

In relation to household victimisation, it found that while security measures to prevent burglary - such as safety latches, outside sensor lights, deadlocks and security chains on doors - were quite common, such measures were not as prevalent as in England and Wales. Moreover, most thought that their houses would be "very" or "fairly" easy to break into.

Most respondents thought that security precautions against burglary were at least to some extent effective. We are not able to say with confidence from our data whether their perceptions in this respect were valid, since we were unable to compare the number or type of security measures which victims and non-victims had in place at a particular time. However, a comparison of the security measures which victims of burglary had in place at the time of the offence with the security measures reported by all households at the time of interview reveals a marked difference between the two. Victims at the time of the offence were substantially less likely than all households at the time of interview to have had virtually every type of security measure, suggesting that security measures do have a preventive effect.

In the light of this, the relatively low levels of security measures taken are a matter of some concern. However, it is the houses of lower socio-economic groups that appear to be most lacking in this respect. This suggests that physical security is not always the panacea for burglary: not everyone can afford effective security measures, and the danger is that good physical security in some houses may simply `displace' the crime to other neighbourhoods or residences which are least able to absorb its impact. The challenge is to develop crime prevention strategies which are accessible and which reduce the risk of crime amongst all social groups.

Women's Safety Survey 1996

Introduction

The National Survey of Crime Victims explored the experience of victimisation of 5000 randomly selected New Zealanders. The Women's Safety Survey aimed more specifically to explore violence against women by their male partners. Thus women who had already participated in the National Survey of Crime Victims and who were currently living with a male partner (referred to in the text as women with current partners), or who had been living with a male partner within the last two years but were not now (referred to in the text as women with recent partners), were invited to become part of a pool from which a sample of 500 was then randomly selected. This eventual sample was made up of 438 women with current partners and 71 women with recent partners (nine women were included in both of these sub-samples as they had both current and recent partners). 351 of the 500 women were non Maori and 149 were Maori. Women were interviewed either by telephone or face to face, depending on their preference.

Objectives of the Women's Safety Survey

The principal objectives of the Women's Safety Survey were: to provide an alternative measure to police statistics of the extent of violence against women by their partners; to provide an alternative measure to the National Survey of Crime Victims of the extent of violence against women by their partners; to describe the context and circumstances of violence against women by their partners; to describe the consequences and effects of violence by women's partners on women and their children; and to identify the people and agencies women who experience violence by their partners talk to or approach for help and describe the women's assessment of that help.

Ethical issues

There are many ethical issues to be considered in surveys of this type: women are being invited to disclose potentially distressing events to comparative strangers and, because of the design of this particular survey, many of the women who participated were still living with their abusive partners and could be placed at risk by participating in the survey. The researchers sought advice about this from agencies working in this area and as a result took a number of steps to ensure as safe an environment as possible: women were allowed to determine the method of contact and time and place of interview best suited to them; interviews were carried out in private; and a toll-free number was given to enable women to call back if they wished. In addition, women were provided with information about relevant services or agencies where this was requested. Interviewers were also given special training and briefing; and support systems were set in place for them.

Key findings

The prevalence of psychological abuse

Women were asked about whether or not they had experienced any of six items of psychological abuse. Examples of these are `he insists on knowing who you are with and where you are' and `he puts you down in a way that makes you feel bad'. More than two fifths of New Zealand women with current partners and the vast majority of New Zealand women with recent partners reported they had experienced at least one type of controlling behaviour. Maori women were significantly more likely than non Maori women to report experiencing controlling behaviour by their partner. The majority of women who had experienced these behaviours gave descriptions of them which were consistent with their partner's wish to exert `power' or `control' over them. Some women also commented that the cumulative effects of psychological abuse were more damaging for them than the effects of physical abuse.

The prevalence of physical abuse

Women were asked about whether or not they had ever experienced any of 22 items of physical or sexual abuse ranging in seriousness from `using a weapon against you' and `choked or tried to choke you' to `threatened to slap you' and `threatened to push or grab you in a way that could hurt you'. Overall, around a quarter of the women with current partners and almost three quarters of the women with recent partners reported that they had experienced at least one act of physical or sexual abuse by their partner. The most common behaviour reported was being `pushed or grabbed in a way that hurt'. Maori women were significantly more likely than non Maori women to report that they had experienced at least one act of physical or sexual abuse by their partner. Two percent of the women with current partners and 22% of the women with recent partners reported that they had experienced ten or more acts of physical or sexual abuse by their partner. Maori women were significantly more likely than non Maori women to report that they had experienced multiple acts of physical or sexual abuse by their partner.

The prevalence of physical abuse within the past 12 months

Women were also asked about whether or not they had experienced any of these same 22 items within the last 12 months. More than a quarter of the Maori women and a tenth of the non Maori women with current partners who participated in the Women's Safety Survey reported experiencing at least one act of physical or sexual abuse in the past 12 months. Again, the most common behaviour reported was being `pushed or grabbed in a way that hurt you'.

Seriousness of physical abuse

Three indicators of the seriousness of abuse were used: medical or hospital treatment, fear that a partner might kill them, and the woman's own assessment of seriousness. One percent of the women with current partners and 8% of the women with recent partners reported that they had been treated or admitted to hospital as a result of their partner's violence; the comparable figures for Maori women were 2% and 19%. One percent of the women with current partners and 7% of the women with recent partners also reported that they had received medical treatment from a doctor as a result of their partner's violence; the comparable figures for Maori women were 3% and 24%. Further, 3% of the women with current partners and 24% of the women with recent partners reported that they had been afraid that their partner might kill them; the comparable figures for Maori women were 5% and 44%. Half of the women with current partners who had experienced violence by their partners and more than three quarters of the women with recent partners who had experienced violence by them rated the violence they had experienced as `very serious' or `quite serious'. Similar proportions of women said the violence they experienced had affected them `very much' or `quite a lot'. The majority of women who believed that their children had witnessed or heard the violence they had experienced at the hands of their partner also believed that it had had an effect on their children.

Coping with partner abuse

A range of strategies are used by women to cope with violence by their partners. A sizeable minority of women with current partners and most women with recent partners reported that they had left their partners for some length of time because of their violent behaviour. `Leave him permanently' was the principal strategy women suggested for making themselves safer. Most of the women who had experienced violence had spoken to someone about it, mainly their partners, their immediate family and their friends and neighbours. Most people or agencies approached by women who have experienced violence are rated `very' or `fairly' helpful by them.

Changing partners' behaviour

Only just over a tenth of the women who disclosed any level of physical violence had ever asked the police to come to their home to deal with their partner's violence. However, more than two thirds of the small number of women whose partner had been dealt with by the police were `very satisfied' or `satisfied' on the last occasion this occurred. The most common influences mentioned by women in changing their partner's behaviour were `fear of losing me' and `counselling'. Neither `police action' nor `court action' featured in the `top five' influences which women believed might change an abusive partner's behaviour.

Methodological caveats

The overall response rate for the Women's Safety Survey was high: 79%. Because of the way in which the sample for the Women's Safety Survey was drawn, however, this response rate has to be viewed within the context of the response rate for the National Survey of Crime Victims. The overall response rate for that survey was 57%. It is possible that the response rate for women was significantly higher than for men and there are plausible reasons for suggesting this. However, we do not know this and the potential impact of the low response rate in the National Survey of Crime Victims has to be kept in mind in extrapolating from the findings of the Women's Safety Survey.

The sample for the Women's Safety Survey was intended to be a national random sample of women currently living with male partners or who had been living with a male partner in the last two years. For the reason just set out, we cannot be sure that it was. In addition, women who reported in the National Survey of Crime Victims that they had experienced partner abuse were slightly more likely than women who did not report such abuse to agree to participate in the Women's Safety Survey. This may mean that the figures presented in this report over-estimate the extent of violence against women by their partners by having within the sample at the outset a greater proportion of women with experience of abuse by partners than one would expect in an entirely random sample. The questions in the National Survey of Crime Victims, however, related to any of the woman's previous partners and not to the specific partner under discussion in the Women's Safety Survey and so again we cannot be sure that any bias has occurred. This possibility also needs to be weighed against the likelihood that some women who participated in the Women's Safety Survey - and, more particularly, some women from certain sub-samples such as older women or women of high socio-economic status - will have under-reported their victimisation for a range of reasons: shame, embarrassment, self-blame and so on. Previous research would certainly support this suggestion.

Further, it is not possible to distinguish within this survey whether any of the differences in women's experience of violence at the hands of their male partners is due to differences in some women's willingness to disclose their experience of this violence, to differences in their actual experience of this violence or to some interaction between the two. In addition, some of the sub-samples referred to are quite small - for example, the number of women in the sample with recent partners. For these reasons, the findings in the body of this report have to be treated with caution. In this area, survey techniques, no matter how carefully designed, are not capable of producing `actual' or `accurate' figures. All that can be hoped for is that the figures and findings produced are `more accurate' or `less biased' than those currently available to agencies and Government and that the knowledge base about violence against women by their partners is thereby extended.

Sector and Agency Responses

In considering their responses to the victimisation survey, the seven sponsoring agencies recognise that there are no simple solutions to the problems identified. They believe that long term strategies are required with inter-agency cooperation and coordination in the development of strategic initiatives and policies for victims of crime. Furthermore, this process will be supported by the continuous measurement of victimisation, through repeat surveys of victims, to determine changes against the baseline findings of the first survey.

The initial responses of the sponsoring agencies, in terms of the specific strategies and policy areas affected by the survey information, are grouped under the following general issues.

Organisational Strategies

The survey findings provide support for organisations to re-affirm, with greater confidence, those strategies that are currently directed at dealing with the causes and effects of victimisation, family violence, community security, anxiety about crime, and victim support. However, there may be a need for some re-assessment of current strategies, policies and targets taking into consideration the findings of the survey.

Responses to Crime

The Responses to Crime Strategy aims to prevent crime and to reduce its impact. This is achieved both by ensuring that advice to Government is consistent, has a strategic focus and provides clear policy choices, and by having a system of explicit agreed sector goals and objectives linked to performance indicators with integrated agency responses to crime.

The survey findings indicate that inter-agency strategies need to target high-risk groups and concentrations of victims, especially those at risk of repeat victimisation and violence. There needs to be a continued focus on dealing with anxiety about crime and being a victim of crime, targeted crime prevention programs and communication strategies regarding personal safety and security in homes and in the community. These will be achieved, in part, through continued partnership initiatives at the local and community levels.

Crime Prevention

The results of the 1996 New Zealand National Survey of Crime Victims are a confirmation of policy actions that the Government has begun to take over the last three years to broaden its approach to the problem of crime in New Zealand. One such policy shift has been to place strong emphasis on crime prevention. In 1996, the Government approved $11 million for programmes and new initiatives designed to prevent family violence. In 1997, the Government allocated $7.15 million for programmes aimed at youth at risk of criminal offending. In the next three years the effects of these programmes will be evaluated for their effectiveness in tackling the problem of crime.

The survey has the potential to contribute to crime prevention in two ways. Firstly, it provides valuable information on targeting the activities of government departments and new initiatives towards the prevention of both re-offending and re-victimisation in respect to both violent and property crime. Since a relatively small number of victims suffer a disproportionate amount of offending against them, a significant impact on the incidence of crime will result if this type of offending is targeted effectively. Secondly, the survey provides a benchmark against which crime prevention, law enforcement and criminal justice effectiveness can be measured in future years. This is important for the ongoing development of effective strategies and policies for the criminal justice sector as a whole.

Family Violence - Male Partner Violence

The survey finding that male partner violence has a high prevalence for New Zealand women, particularly for Maori women, is an important addition to the growing information base on family violence. The Women's Safety Survey findings on the significance and effects of partner violence will be used in evaluating policy and services across a range of Government agencies.

The survey findings give further impetus to work relating to implementing and evaluating the provisions of the Domestic Violence Act 1995, the Police Family Violence Policy, the Family Violence Research Priorities Project, the Maori Family Violence Prevention Services Contestable Fund, and the development of Family Violence Protocols for the Health sector.

Women

The survey findings will assist the Ministry of Women's Affairs and other agencies to evaluate existing policies and services, and to develop future initiatives to combat the victimisation of women. Again, the survey finding that male partner violence has a high prevalence for New Zealand women, particularly for Maori women, is an important addition to the growing information base on family violence.

Responses to Maori

The survey findings will assist Te Puni Kokiri to work with other agencies to improve the agencies' responsiveness to Maori and will inform the projects on Responses to Offending by Maori, and the Police Responsiveness to Maori Strategy.

Victim Support

The needs of victims and the provision of effective support for victims will be informed by the survey findings. The survey information will inform the Coalition Agreement initiative to review Victim Support. Other areas affected should include training for agencies' personnel and targeted education programmes for the community both generally and for at risk groups.

Criminal justice system

Within the criminal justice sector, the survey findings will be significant for the evaluation and development of strategies with respect to quality service delivery, the needs of victims (ie to be kept informed and dealt with appropriately and consistently), and the training of personnel. This includes mechanisms for improving performance and service to those few groups of victims who expressed dissatisfaction in their dealings with the criminal justice sector.

Young People

Inter-agency projects focusing on `youth at risk', which are concerned mainly with young offenders, will be informed by the survey findings in relation to the risk of young people becoming victims of crime, and their anxiety about being victims. Those working with Maori offenders and `victim needs' will also be informed by the findings of the survey. Strategies promoting safety among young people, including those developed by Safer Community Councils and in schools should also incorporate the survey findings.

Responses to Pacific Island People

The survey findings are significant for Pacific Island people and will inform the Offending by Pacific Peoples Project being lead by the Ministry of Justice. However, the survey also indicates a clear need for further research in this area and inter-agency strategies need to take due cognisance of this.

How to obtain a copy of these reports

These reports can be bought by contacting:

or,

Freefax 0800 804 454

The reports are also available from Bennetts Bookshops around New Zealand.