NZ Crime and Victims Survey - video transcript

Dr Michael Slyuzberg:

Thank you (inaudible) and thank you Oranga Tamariki for inviting us.  It's kind of mission impossible to report the results of a crime and victims survey within 20 minutes so my first suggestion is don't pay too much attention to numbers.  I will bombard you with numbers but it's more about story rather than numbers and all the numbers are available from the website in many different formats. 

So, New Zealand Crime and Victims Survey, a national survey which substitutes the formerly run New Zealand Crime and Safety Survey, which was disconnected in 2014.  We believe that NZCVS is really a big improvement, both in terms of methodology and in terms of value for money, and in terms of effectiveness and in terms of accuracy which is probably the most important.  The survey is delivering all core crime and victimisation measures.  For New Zealand what is important that we are talking about both reported and not reported to police crime, so this data is very, very different from admin data provided by police.  The survey covers 12 months back from the date of an interview, that's another important thing to understand.  The results we report is not crime in 2017, it's crime 12 months before the date of interview and the interviews are ongoing. 

We interview 8,000 people annually and meanwhile we have money for three years, then we will see.  The case bit for Treasury is on my desk but just keeping finger's crossed.  However, 24,000 interviews will be finalised by September 2020 and it's a lot.  We are using two samples about two-thirds it's a core sample developed and provided by Statistics New Zealand, about one-third is Māori boosting sample based on data we received from Electoral Committee.  These are face to face interviews, no way to make it on phone or online, it's too sensitive and too sophisticated.  The important difference with NZCASS is that we allow people, especially highly victimised people, to cluster, to group similar incidents, which is very important in terms of resulting data and accuracy because you can imagine if a person says that they were victimised, family violence types of incidents, 120 times a year we can't ask to describe all these incidents and NZCASS in this situation imputed 119 based on 1, which was far from ideal and now we have I believe much, much more robust approach. 

The survey consists of two modules, one module is the core one which is repeated without any changes every year which allows us to build robust time series, the second module is what we call in depth module and it is changing every year and focussed on particular prioritised types of offences and for the first year it was family violence.  And the last very important difference is that our coding system of four types of offences was developed by our friends from police, and so we are pretty confident that our coding and police coding is very similar.  We try to be as transparent as possible in terms of the results and every year we are going to release the whole set of reports, including top line report, methodology for each year, which will be consistent but slightly different because of in depth modules, key findings reports, and the whole set of topical reports on prioritised topics which we are working right now, and plan to release by Christmas many of them.  All of this is available from the website.

Now, what did we find?  71% of adults, adults means 15 years old and older, did not experience any crime over last 12 months.  How to interpret, well, it it's up to you.  It's a significant majority but it means that 29% did experience which is not really good.  80% of households didn't experience any household crimes, four out of five.  The most widespread offence is burglary which was not surprising, then it was followed by harassment and threatening behaviour and then fraud and deception.  We returned fraud and deception as an offence type to Crime and Victim Survey; it was not part of the survey in 2014 and 2009.  Māori are significantly more likely to be victims.  On all these graphs orange means statistically significant difference from national average.  Interestingly, Chinese people are significantly less likely to be victims.  Older people are more safe than national average while the highest risk is for 20-29 years old group.  One important thing, the survey covers only residential households.  We don't ask people in institutions, we don't ask people in prison, so this may a bit affect all the people who live in retirement villages, for example. 

There is no difference overall between the level of victimisation for men and women.  It's perfectly equal, 29.3%-29.4%, but when you will start breaking down these numbers by offence types you will see in a few minutes the picture is hugely different.  There is very little difference in crime between major urban centres, and you can see that we live in the crime capital of New Zealand, 33%, but it's not significantly different from all other urban centres.  And again you will see that if you will go down to suburbs level it will change. 

Some more important findings; very clear correlation, very clear dependence between the level of victimisation and their life satisfaction.  Perfectly linear with very high correlation coefficient.  I am not talking about causality(?), I am not saying that happy people experience less crime, maybe it's the other way around, maybe it's those people who experience less crime are more happy, but still there is a very clear correlation.  Also a very clear correlation between the level of victimisation and perceived safety of people, which is not surprising of course.  Surprising is that we didn't find any difference between disabled and not disabled people.  We expected it may be higher for disabled people; no, it's not.  And when we go down to offence types, grouped offence types, it's also not.  However, there is a huge difference in victimisation level for people with high mental health pressure.  We look at three grades, low, moderate and high, even for moderate mental pressure the level of victimisation is significantly higher.  Not surprisingly, the higher is the area depravation where people live the higher is level of victimisation, especially for household crime.  People with high level of financial hardship have more chance to be victimised, and those living in larger households, and larger households are five or more people, more likely to experience crime.  Now, this is a very interesting graph; 4% of victims report 50% of incidents.  It's a hugely high concentration.  And 11% of victims, one tenth, report three-quarters of all incidents, and we are currently working on a topical report, who are these 4%, what is the difference between them and other people, whom we should focus on in the first instance. 

Now family violence and sexual violence.  Family violence for the purpose of the survey is any interpersonal violence and some property violence, which is done by family members.  Family members in a very wide sense, we often -- we always say family/whanau, so it's more than just immediate relatives.  This includes intimate partners of course, both current and ex-partners.  I said that the survey asks about what has happened 12 months before the interview, there are 2 exceptions.  We ask about sexual violence experience and intimate partner violence experience over their lifetime as well, and the numbers -- the lifetime numbers are really shocking.  One in four adults experience sexual violence of different kinds during their lifetime, one in three women.  One in six adults experience intimate partner violence over a lifetime, one in five women.  Back to 12 months, surprisingly the number of adults experiencing family violence from their partners is more or less equal to violence from other family members.  I expected this proportion will be higher but it's not.  71% of victims are women, 71%.  Māori people experience family violence twice more often than pākehā.  In terms of age, 40% are young people, 15 to 29.  Interestingly, and they can't explain it, we have a bit of a drop for the age group 30 to 39 and then it goes up again for 40 to 49 and 50 plus.  I don't know how to explain this drop and maybe the second year will clarify -- the second year data collection will clarify the situation.  We assess that almost 45,000 of adult New Zealanders victimised by their intimate partners, 30,000 current partners, 16,000 ex-partners.  And for intimate partner violence 77% of victims are women, almost half between 15 and 29 years old, and Māori people experience almost three times more intimate partner violence incidents than the national average. 

Now we analysed psychological violence and found that more than 100,000 adults experience different types of psychological violence over 12 months which included different things starting from controlling behaviour of different kinds and up to pressing into paid work and stopping from doing paid work.  Here the gender proportion is more balanced, 54% of victims are women.  If you will combine physical and psychological violence then the number of victims achieves 160,000 adults. 

Sexual violence, almost 200,000 sexual assault incidents with 90,000 victims.  Here the proportion of women is 80% of victims and 66% of victims are 15 to 29 years old.  However, from ethnical perspective there is not much difference between Māori and pākehā for sexual violence.  We tried to analyse what are the immediate drivers of crime and of course it's perceptional data, it's what victims told us what they believe triggered the incidents.  We found that about 20%, one of five victims, believe that the incident happened because of offender's attitudes towards particular things, such as race, ethnicity, sex, age, religion, disability and so on.  Sex is the obvious leader in this list.  All other reasons, all other triggers are on the 7% and less percentage level.  Sex is their major perceived driver for one-third of interpersonal violence.  People believe that one of three incidents of current partner violence was triggered by financial issues.  44%, almost half of family violence incidents, triggered by arguments and it is followed by jealousy.  And 40% of family violence incidents happened when an offender was under the influence of alcohol or drugs, 40%.

And finally we analysed within our in depth module family violence victims' experience, what do they feel about support they received after the incident happened.  And it's interesting and to me a quite concerning picture.  A huge majority, 90% plus, know about support organisations, about helplines, about phone numbers, virtually everybody.  But only 23% used this.  And when we asked why not people -- about 30% of those who did not use support organisations said, "No, we don't need help" and then they said, "It's a private matter and we wanted to handle it ourselves" which is great if they are able to do it but in many cases this results in repeated victimisation, in increased level of violence, and I think one of the obvious action points from the results of this survey is that we need to talk more with victims and we need to explain to them why it is good to ask for help and why it is not a shame to ask for help. 

Just about 23%, under a quarter of all crime, was reported to police.  The percentage is very much varying depending on types of crime.  82% of theft from vehicles reported to police which is not surprising, it's insurance, they have to.  On the other hand, only 7% of fraud which is probably also not surprising because in many cases people report fraud to their banks, to their financial companies, to their insurance companies and not to police.  I was surprised with the pretty high level of family violence reporting, it was 37%, compared to burglaries 36%, or the same, which is to me a very positive sign.  And interestingly victims more often report, as Bronwyn said, their intimate partner violence than violence from other family members.  My hypothesis is that intimate partner violence is much more repeatable and victims just lose their patience, while other family members it may be more episodic, less frequent. 

There are a few factors which affect reporting to the police.  We asked people, "Do you consider what has happened as crime or just something wrong but not crime?" and those who don't consider it crime report it to police even in 9% of cases, while those who see it as crime in 33% of cases.  And we also asked people to categorise the seriousness of crime on a scale from zero to ten.  So those who assess the incident as serious, starting from level seven out of ten, their level of reporting to police is around 40%. 

Finally, we tried to use the survey data as much as possible to support policy development, operational decisions, and whatever management decisions.  Because of this we try to share this information as wide as possible.  From September - and I am keeping fingers crossed - but from September we hope our database will be available from IDI(?).  Disclaimer, 93% of the database, because we received consent from victims for 93% of responses.  Stats is using our victimisation indicators for their Indicators Aotearoa project.  We continue developing our website and we hope in some time to be able to offer interactive tools for people to explore more about that survey data.  And one thing I would like to urge you when you will use CVS data, please don't compare the results with NZCASS or internationally because in many cases and most cases we are talking about very small differences, and these are different surveys, this is different methodology, this is different sample size, this is different coding system, so they are just incomparable and the comparison may result in very misleading conclusions.  Where to find, as I said, everything is on our website, we have there multiple formats starting from 150 pages full size survey and up to a set of A3 infographics, and up to Excel spreadsheets with aggregated data.  And you are very welcome to contact us and ask any questions you have about the survey.  Thank you very much.