Evidence Centre Seminar: August 2021
Published: August 20, 2021
The seminar featured two presentations, both with a youth justice focus. We learned more about the NZ Crime and Victims Survey with particular reference to young people from two Ministry of Justice presenters. The Evidence Centre reported on the pathways for young people that can lead to involvement in the youth justice system.
Offences by family members and victimisation of young adults
Findings from the NZ Crime and Victims Survey
Dr Tadhg Daly and Kate Preston from the Ministry of Justice spoke about the methodology used in the NZ Crime and Victims Survey (NZCVS) and shared findings from both the latest survey and pooled data from the last three survey years.
A key finding was that younger people, aged 15–29 are significantly more likely to be victims of personal offences and interpersonal violence offences in the last 12 months.
One in four females and one in nine males aged between 15 and 19 had experienced sexual assault in their lifetime. However, this age group also reported relatively high feelings of safety.
For more detail about the NZCVS, including full reports and data tables, see the Ministry of Justice website: NZCVS resources and results | New Zealand Ministry of Justice
Ministry of Justice seminar: NZ Crime & Victims Survey findings in relation to young people
DR TADHG DALY: Kia ora koutou, thank you very much for having us here, so today me and Kate are going to talk about some results from the New Zealand Crime and Victim Survey, specifically focusing on offences by family members and young adults. To give you a brief outline of what we're going to talk about, I'm going to give you a brief overview of what the New Zealand Crime and Victim Survey is and does, then I'm going to give you some sort of high-level results about offences against young adults, and then Kate is going to talk you through some of our data on offences by family members and give you a little bit of a rundown on our future plans.
So why do we need a Crime and Victim Survey? The main reason we need a Crime and Victim Survey is that 75% of crime incidents are not reported to the Police and so that means that Police admin data is really only focusing on about 25% of crime incidents. The other thing that a Crime and Victim Survey does is, for Police admin data, it focuses a lot on offenders, so, with a survey, we are able to catch or collect a richer source of information about victims' experiences and perceptions and compare that to non-victims'.
A brief rundown of the survey. We talk to 8,000 people aged 15 plus each year. It's a national survey, randomly selected households across the country using Stats New Zealand mesh blocks. We generally get an 80% response rate. The surveys are face to face using CAPI and CASI, meaning some questions are asked by the interviewer and some more sensitive questions are filled out on an iPad by the respondent. So far, we have three years of data starting from 2018, and the fourth cycle is currently in the field.
So when we go to people's houses and we ask about what's happened to them in terms of victimisation, we cover the last 12 months up to the time of that interview, and we don't ask them, "Has a burglary happened to you, has an assault happened to you?" we ask about scenarios, so we say, "Has somebody been on your property without your permission and taken something, has somebody hurt you in any way?" and then what we do, if they say yes to any of those screeners, is that we get a description of that incident and then we code that incident as an offence or not an offence in line with Police protocols, and then we also ask whether that offence was reported to the Police or not, or whether that incident was reported to the Police or not.
Some features of the survey is that we have a Māori booster sample to make the survey more representative for Māori. We let people bundle incidents together. So on previous surveys, people weren't allowed to cluster incidents together; in this survey, if people had a lot of things happen to them, we let them bundle them into a cluster victim form so they can talk about a lot of incidents together, and that just lets us collect more information about what's happened to someone in the last year.
Each year we have an in-depth module which changes from year to year, so, in the second year, it was about wellbeing and perceptions of the criminal justice system, and last year it was about help seeking for victims of family violence. And all our coding is supported by New Zealand Police so that we know that what we are calling an offence is the same thing that Police are calling an offence.
So looking at sort of high-level victimisation over the last three years, these are the incidents and prevalence rates. On the high level, we haven't seen a big change in victimisation, so about 30% of the population experiences an offence each year. We have, after COVID-19, seen a drop in household offences, and, specifically, burglary, but that drop hasn't had enough of an impact to affect the higher level offences.
So now I'm just going to give you sort of an overview of some high-level sort of data or results that we have on 15 to 19 year olds. In some case, I'll also talk about 20 to 29 year olds, just as the next oldest group, and sometimes I'll be combining those two groups together. Because the data, the victimisation rate hasn't changed very much over the last three years, we'll be mostly focusing on pooled data, so combining our three years of data together to give us more information.
So this is our overall rate of victimisation for personal offences by age. So these are prevalence rates, and I should point out that the orange bars mean statistically significant difference from the New Zealand average. So what this graph is showing is that 20% of 15 to 19 year olds experienced a personal offence within the last 12 months, compared to 15% for the New Zealand average, and you'll see that, as people get older, there's a downward trend, people are less likely to get victimised, with 65 years and older, 9% experiencing an offence within the last 12 months.
This is a terrible graph but I'm going to sort of talk you through it. If you look from left to right, what I've done here is, basically, just comparing that 15 to 19 year old group to the New Zealand average by offence type. So, on the far left, we have interpersonal violence where 15 to 19 year olds are almost twice as likely to experience interpersonal violent offences. Looking at burglary, 15 to 19 year olds are not that much more likely than the New Zealand average to experience a burglary, more likely to experience theft and damage offences, not more likely to experience vehicle offences, and so you can imagine with these property offences, as you're younger you have less property so you're probably less likely to experience property offences.
Fraud and cybercrime, actually significantly less likely to experience fraud and cybercrime offences, not sure why that is, it may be a tech-savvy type thing. And finally, sexual assault, so over double, over twice as likely to experience sexual assault within a 12-month period. And finally, at the end here, we have "highly victimised", so it's not quite significant but that 15 to 19 year old age bracket is more likely to be highly victimised, and what we mean by highly victimised is someone who has experienced four or more offences within a 12-month period.
So, if we look at this a little bit by demographics, for this age group, females of the 15 to 19 year old age group are more likely to experience personal offences compared to the New Zealand average versus males of this age group. They're still both high, and the error bars are quite high when we're breaking the groups down by this much, but the trend is just for females to be a bit higher.
When we look at disabilities, we see quite stark differences by age, so when we look at disability overall, disabled versus non-disabled, we don't actually see a difference in victimisation, but, when we break it down by age, we actually see quite huge differences. The reason for this is that the disabled population tend to be older and older people tend to be less likely to be victimised than the average, so, when you do break it down by age, you see these quite huge differences for young adults with disabilities, so 50% experiencing crime within a 12-month period compared to 30% average, New Zealand average.
For those age groups, broken down by ethnicity we see Māori, 15 to 19 and 20 to 29, more likely to experience any type of offence, any victimisation, and the same for New Zealand European, 20 to 29 years old, but we don't get the same Pacific and Asian, more commensurate with the New Zealand average for those age groups.
I don't know if I mentioned this before, but we have a couple of questions in the survey that are not about the 12-month period, they're about the lifetime, so we ask about lifetime sexual assault and we ask about lifetime intimate partner violence. And so we get quite stark results for young people on sexual assault, so 2 in 10 adults, about 18% in that 15 to 19 year age range had been subject to sexual assault in their lifetime. 1 in 4, when you break that down by sex, it's 1 in 4 females, or 27%, and 1 in 9 males, 12%.
So despite these sort of higher rates of victimisation for this younger age group, young people aren't typically less trusting than other age groups, so when we ask them, "How often do you expect most people to take advantage?" they're very similar to the other groups in how much they say "a little of the time" or "none of the time". Young people, 15 to 19 years old, also tend to feel a lot safer, paradoxically, than other age groups, so we've got about 65% of 15 to 19 year olds are rating their feelings of safety as a 9 or a 10 out of 10, and the next highest group there is the actual safest group, which is the 65 years plus group, and that's only 54%, so that might be showing us sort of a feeling of invincibility that young people tend to have.
Finally, when we ask people of this age group if they saw a crime incident, how likely would they be to report it, 15 to 19 year olds are the least likely to say that they'd be very likely to report an incident that they witnessed, so 58% said they would be very likely to report an incident, all the other age groups are a lot higher, starting at 68% for the 20 to 29 year group.
And when you compare this to what actually happens in terms of reporting rates, true to their word, younger people are less likely to report offences that happen to them, so we're only seeing about 17% of incidents being reported to Police by 15 to 19 year olds versus that 26% reporting rate average.
So that's basically some high-level results that we have on young people from the survey; now I'm going to pass it on to Kate Preston to talk about some of our data on offences by family members.
KATE PRESTON: Thank you. Kia ora kotou, it's great to be here today. As Tadhg said, I'm going to walk you through some of our results about offences by family members, and I'm also going to be focusing on our pooled data that's using three years of data combined together because we don't see much change in the high-level rates of these offences over time and we are working with quite small sample sizes for these types of offences.
So more than a quarter of interpersonal violence offences, actually around a third, were by family members. When I talk about interpersonal violence offences, I'm covering physical violence, sexual assault, threats and harassment, and property damage.
So we see here that about 7% of New Zealanders had an interpersonal violence offence over the last 12 months, and within that around 2% had an offence by a family member, and about 5% by someone else they knew or a stranger. We have a definition in the NZCVS for offences by family members which is very similar to what I just described for interpersonal violence, we just treat some of the property damage offences slightly differently. So it includes episodes of these kinds of offences in the last 12 months where the perpetrator was a partner, ex-partner, or other family or whānau member of the victim.
This is a very offence-based measure so this is why we don't call it family violence in our reporting, because family violence encompasses a much wider range of patterns of behaviour that cause harm or can coerce or control the person. The survey takes a very incident-based approach and collects about specific episodes of these offences.
So around 1 in 50, or 2%, of adults had been affected by one of these offences by a family member in the last 12 months, that's about 90,000 adults, and together they experienced around 240,000 offences. So they make up about 7% of all victims in our survey and about 14% of all offences, showing that they experience kind of a disproportionate burden of victimisations.
Around three quarters of the offences against family members were by intimate partners, the rest being by other family and whānau members. Intimate partners include current partners and ex-partners. And about 4 in 10 of the offences were physical assaults or robbery, or what we'd group together as physical violence, 3 in 10 were threats and harassment, 2 in 10 were sexual assaults, and 1 was property damage. We get asked quite a lot about the overlap of offences by family members and sexual assault, so you'll see here that, as I said, 2 in 10, or 20%, of offences by family members were sexual assaults, and, similarly, around 25% of offences by family members were sexual assaults. Now, that's mostly being perpetrated by intimate partners where it's family members perpetrating a sexual assault, and, overall, more than half of sexual assaults are by someone that the victim already knew before it happened.
Offences by family members are highly gendered, I'm sure many of you will be aware of that already. Women were three times as likely as men to experience an offence by family members in the last 12 months, they were four times as likely to experience an offence by an intimate partner, I think 75% or 70% of offences were perpetrated by men against women.
And all population groups are at risk, so men, women, all of our groups shows rates of offences by family members that we look at, but some really carry a disproportionate burden of this harm. A particularly high-risk group are people who have the marital status of being separated. This is at the time of the interview, not necessarily at the time of the offending, they could have separated since the offending happened. So, compare to 2% of adults in the population, around 11% of them were victims. I think for young, 15 to 29 year olds, that increases to about 24%. Bear in mind some high confidence intervals for some of these small groups. Single-parent households also at high risk, households with four or more children, those living in Government rental accommodation. Other high-risk groups included Māori women, sort of any interactions of these things with women, very high-risk groups.
Adults with disabilities were three times as likely as other adults to have been a victim of offences by family members once we account for age. So, as Tadhg said, our older population is at lower risk of most offending, but, if we imagined the disabled population having the same age structure as the overall population, we would expect them to be at three times higher risk than the average New Zealander.
Just over half of all the adult victims lived in households with children, so I think that's around 60,000 or so of our victims were living in households with children. Women were at higher risk of being a victim if they lived with children, about twice the average risk, whereas, for men, there was no difference by whether or not they were living with children.
15 to 19 years old are at sort of slightly higher risk than average of being a victim of offences by family members out of all adults, but that's not a statistically significant difference, at 3% compared with 2% in the overall population. Our 20 to 29 year old group is at significantly higher risk, I think their rate is around 4%, and then it drops off again as people get older. This graph just shows sort of the make-up of victims, so about 7,000 were 15 to 19 year olds making up about 8% of all adult victims, so, despite the lower risk for our 30 to 64 year old group, they make up the bulk of our victims because they're such a large group in our population.
Tadhg mentioned we have some questions on lifetime rates as opposed to 12-month victimisation. This picture here talks about lifetime intimate partner violence, either being experiencing force or violence by a partner or ex-partner, or threats of force or violence by a partner or ex-partner. Again, we see very high rates for women in their lifetime; also, a lot of men have experienced intimate partner violence in their lifetime. Some other high-risk groups are people with diverse sexualities, highlighting that intimate partner violence doesn't just occur in same-sex relationships.
About a third of offences by family members, going back to our 12-month measure now, are reported to Police, that's a little bit higher than the average offence. There was no difference by whether the perpetrator was a partner or another family member, and common reasons given for not reporting were that it was a private matter, a personal or family or whānau matter, or that it was too trivial, or there was no loss or damage, or it wasn't worth reporting. I think shame and humiliation also come up as quite common reasons for these offence types too.
This result is only from our last year of data because it's a new question. One in six adults knew of someone else who had experienced a family or whānau incident in the last 12 months, and about 55% of those then went on to have further involvement with the person they knew. So we're going to do a little bit more exploration of some other questions we added about barriers to helping those people, or what kind of involvement they had, but I think this highlights a point of help for many of our victims, and we need to be able to resource our communities to support those who are experiencing harm.
We have lots of other information in the survey on victims' experiences, how much, how often alcohol or drugs were involved in the offence, factors related, so we commonly hear that separation or arguments were factors in these offences, the types of emotional reactions, depression and anxiety, quite a lot of our victims said they experienced these as a result of the offences. We have information on whether they were injured, whether they sought medical treatment, and whether they perceived the incident to be a crime or not and how serious they thought it was. These are available on our data tables and reports online but I don't have time to go through these all today.
I'm going to just finish up highlighting some of our reports that are available on our website. Our Cycle 3 report was published in June and has lots of information, including a chapter on family and sexual violence and lots of information about our young populations. Tadhg produced a report on Māori victimisation that is up on the website too, and we have a report on victims' perceptions of the criminal justice system where we showed that victims of interpersonal violence, sexual violence, and offences by family members have some of the lowest perceptions of the criminal justice system in our population.
We're still collecting data right now and we'll be reporting on that in the next year. That includes some new questions on controlling behaviours, trying to look out beyond the scope of offences in family violence, and we're working with the New Zealand Police in our current Cycle 4 with a new module on public perceptions of the Police.
And our data is in the IDI, so that's the integrated data infrastructure linked into administrative datasets, and we'd encourage you all to consider research projects you might like to use this data for and connect it to your own administrative datasets, and we welcome you to have a chat with us about how we can help you to do that, so look forward to hearing from any of you about that, and that's all I have for you today, so thank you, nga mihi for having us today.
Youth Justice Pathways
An examination of wellbeing indicators and outcomes for young people involved with youth justice
Evidence Centre senior analyst Sarah Richardson’s presentation focused on the routes into the justice system for young people – especially the care and protection – youth justice – prison pathway.
She reported that these young people deal with complex factors including unstable home lives, low engagement with education and mental health and substance use issues. Young people with a combination of care and protection experience and youth justice involvement generally fared worse than rangatahi with only care and protection or youth justice involvement.
For background on a range of wellbeing indicators and outcomes for this group, see the full report:
SARAH RICHARDSON: Tēnā kotou katoa. Ke Oranga Tamariki ahau e mahi ana. Ko Sarah Richardson tōku ingoa. Tēnā kotou katoa.
Thank you, everyone. As Rosie has already helpfully explained, today I'm going to talk to you about a piece of analysis that we've done in the Evidence Centre looking at pathways to youth justice and what some of the wellbeing indicators and outcomes look like for young people that go through youth justice compared to those that don't.
So, I thought it would be helpful to start off and tell you a little bit about why we've done this analysis. There's many reasons why you would want to look at pathways through youth justice but one of the driving reasons we did this was that there is a commonly held perception that the care and protection system churns out young people into youth justice which, in turn, churns out adults who go on to offend and end up in prison. While that is the reality for a small number of people, in reality, it's a lot more complicated.
So, as we were scoping this up, we thought, not only do we want to answer that question, but we want to look at what these people look like at three different points in their life, so we want to know what's happening for these people in terms of their wellbeing and underlying factors before they enter youth justice. We want to know what they look like in terms of those things at around age 17 when they're aging out of the remit of youth justice and Oranga Tamariki and then we want to know what happens to them when they leave us. Do they go on and end up serving a prison sentence, do they stop offending, and what does their wellbeing look like?
So, the only way that we could do this analysis was by using the IDI which Kate has very helpfully explained for me what the IDI is, so we pulled out a bunch of indicators from data in the IDI around education, benefits, status, housing, health, for people born in New Zealand between 1993 and 2002 and this meant that we could look at them from childhood and follow them through into early adulthood.
For this cohort, we cut them into -- that sounds awful -- we segmented them into four groups and we did this based on their statutory involvement with Oranga Tamariki. So, when I say statutory involvement, for youth justice that means that they've got as far -- at least as far in youth justice as having a youth justice Family Group Conference and for care and protection I mean that they've got as far as having a Care and Protection Family Group Conference.
So, we have four groups. The first group here is our crossover group, and the crossover group are those people that have had statutory involvement in both care and protection and youth justice. They make up a very small proportion of this cohort.
The second group we have is the youth justice only group and those are the people that have only had statutory involvement with youth justice, no stat involvement with care and protection. So, these are our two youth justice groups. As a comparison, we also looked at people who had had only care and protection statutory involvement and then also the rest of that cohort which is by far the largest group, they've never had a statutory involvement with Oranga Tamariki.
In terms of what these groups look like demographically, our youth justice only groups, which won't be a surprise to a lot of you, Māori are overrepresented as they are in care and protection but more so for youth justice and in terms of gender, these youth justice groups, the majority of them are male. What was quite interesting about this, though, is that in this crossover group where there is a little bit more complexity, there is a higher proportion of females than those that just go through youth justice.
So, in the slides I'm going to show you, there is a lot of detail. Don't worry about focusing in on it too much, there is a paper published online, so if there's anything that catches your eye, you can go back and look at that, so if I gloss over something, it's all available for you.
So the first part of the analysis was looking at what happens for people before they enter youth justice, and for this first part, it's just focused on those two youth justice groups because obviously the others did not get involved in youth justice. So what this really showed us is that young people involved in youth justice have a range of complex factors and things happening for them before they even get there, so there's a lot of opportunity for early engagement with them to perhaps take them down a different path, but what we found in the years before our young person's first YJ FGC is that about 80% of them had had a report of concern to Oranga Tamariki which didn't go as far as an FGC, obviously, around half had been truant from school at some point in their lives and their rates of mental health and substance use treatment were higher in that first, that year before they had their YJ FGC.
An important thing to consider when looking at these as well is that it's not one of these factors in isolation that means someone is going to have interaction with youth justice, it's the combination of all these things together and all the other things they have going on that we can't measure like their cultural connectedness and social bonds, identity, all those things that form a larger picture.
But for the things we can measure, I'll show you some of those. So, I'm going to show you some journey diagrams, again, don't try and digest it all, these are all online, but these show the proportion of young people that had care and protection interactions and family violence notifications leading up to their first Youth Justice FGC. So, these little people on the side, that represents the age they were when they had their first YJ FGC. These are their indicators down here, so for example, we look at 16 year olds who had their first YJ FGC at 16, 7% of those people had a Care and Protection Out of Home Placement in the year before that FGC and then 6% in the one to two years before and 8% in the two to five years before.
What we generally found is that the proportions vary across age, but they tend to be higher for people that have their first interaction at a younger age, and if you look at, for example, the family violence notifications, people that had their first YJ FGC at age 13, 50% of those had had a family violence notification in the year before and then as you go up these age ranges, it decreases a little bit.
So we also have this information for education, so, again, it varies by age, but broadly the picture is here that truancy, stand down and suspension spells kind of increase in frequency as you are leading up to the first YJ FGC suggesting that you're seeing a pattern of behaviour and nonattendance before they even start to address offending behaviour.
Again, mental health and substance use treatment is quite high in the years before the first YJ FGC and varies by age. What's important to note about these mental health and substance use treatment indicators is that they are not a measure of prevalence of mental health issues in the population, they are just that an issue has been identified and those people have had some kind of treatment whether it's cognitive behavioural therapy or, like, a pharmaceutical treatment. And some of that may have come out of addressing offending behaviour or other behaviours that have been seen before they get to YJ FGC.
So that gives you an idea of what these YJ kids look like before they get to YJ, then we looked at what is happening for them in their lives at 17 as they're about to leave us and go off into the adult justice system, so we're back to our four groups now, because we've got some comparatives, and unsurprisingly, the crossover group seems to feel worse across than the YJ only group, so the crossover group were more likely to have committed an offence in the previous 12 months, more likely to have been in a youth justice placement, like a residence than those that had only had youth justice experience, they're more likely to be subject to be a report of concern to Oranga Tamariki, they're more likely to have been in a care and protection placement and they had generally low wellbeing across education, income, health, and they were more likely to experience victimisation. So a whole lot of complex factors happening there.
So, this just shows you the proportion of people in each group who have committed an offence in the year before turning 17. So you can see from this that being involved in YJ obviously was a huge factor but the crossover group were more likely to commit an offence than the YJ only group in the previous 12 months, whether it was a low level offence or a high level offence.
I'm going to show you some of these tables, don't worry about the numbers so much, what we're looking at here is the shading in these tables, so we've got our groups across the top, the red shading indicates that they're faring worse than the other groups and the blue means they're faring better. So, what you'll see across the care and protection indicators is that the crossover group is faring a lot worse than the YJ only group.
In terms of socioeconomic wellbeing, the crossover group were far more likely to have had residential changes in the previous 12 months and they were more likely to live in a low decile area, so that kind of just signifies for us kind of housing instability and instability in their situation which might make it a bit harder to deal with some of the underlying issues that they're having.
Education indicators, we looked at at age 17 as well, so again you can see that that red is highly skewed towards the crossover group, although what is interesting about the education indicators is that if you look at the proportion of these YJ groups who were enrolled in school in the previous year, they're very similar for youth justice, and then when you come down to those that are enrolled in a tertiary course excluding bachelor level, they're actually higher than for our care and protection and no involvement groups so that's suggesting that even though they haven't been in high school, they have been in some form of education like polytech or some NZQA course. Again, this is a measure of being enrolled, it doesn't tell you anything about achievement or how long they were in it for.
Also, a higher likelihood that they've been truant from school in the previous year and for those YJ groups, again, it's much higher than for just care and protection and the general population.
I'll just gloss over that one quickly - so health indicators, these mental health and substance use treatment rates are a lot higher for these groups for the reasons I discussed earlier, that they may have been identified as part of addressing offending, one interesting one here is that the YJ groups, particularly the crossover group, have a higher rate of emergency department admissions in the previous year and this could be because of risk taking behaviour associated with offending or it could also be that because it's higher for CP only as well that sometimes people just can't afford to go to the GP and therefore they present at an emergency department.
One of the most staggering stats, I think, in this thing is that at age 17, just over half of that crossover group have experienced a family violence incident in their immediate family which, when you look at it compared to -- oh, sorry -- the no involvement group is far higher and even far higher than those that have just been involved in youth justice or just in care and protection.
Cool, so you've seen what it looked like before they came into youth justice, what it looks like as they're aging out and now we're going to look at what it looks like for these people between the ages of 17 and 21 once they've left the remit of Oranga Tamariki, and this is the part that people get the most excited about because it's like, are we creating this pathway to prison?
So, again, I kind of use the crossover group as the comparison group because they seem to be the ones that have the worst results and are the most complex. Our crossover group, by age 22, were more likely to commit an offence and they offended at a higher rate. They were more likely to receive a sentence administered by corrections, so a community sentence like home detention, intensive supervision, supervision or a prison sentence, they were more likely to receive multiple sentences, and this was followed in order kind of by the youth justice only group, care and protection only and then the no involvement.
If we look at offending indicators for these groups between 17 and 21, you can see that obviously those that were involved in YJ at all were far more likely to commit an offence in early adulthood than those that weren't, it was 85% for the crossover group and 80% for YJ only, so still very likely to commit an offence. When you look at the average number of offences committed, it's higher for the crossover group followed by the YJ only group and then CP only which is much lower than our YJ groups there.
So this one is showing you the proportion of each group that received a sentence administered by corrections, so the orange here is the proportion that received a prison sentence by the age of 22, the green is a community sentence and the grey is no sentence. Now, these people, even though they didn't receive a sentence from corrections, they may have been to court and received a fine, but because it's not administered by corrections, it's not counted in here.
What the massive difference is here that you can see is that this crossover group, a quarter of them received a prison sentence compared to only 11% of that YJ only group which is quite a big difference, there wasn't too much difference in terms of community sentences, but prison was the big one, and another interesting thing to come out of this was that those people that only had interaction with care and protection and no YJ involvement -- actually, only 2% of them received a prison sentence which goes to show that that misconception that the care system kind of churns out kids into prison is untrue.
This is just a view of that same graph except it's split by ethnicity just to show you what it looks like for Māori, so the outcomes for Māori in terms of prison which won't be a surprise to many of you working in the criminal justice system, is higher, 27% of Māori in that crossover group received a prison sentence compared with 18% of non Māori and it kind of graduates down as you get down to the no involvement group there.
Again, this is just a view of the same graph, it just shows you the number of sentences that these groups are receiving, so the red is the most serious sentence they received is a prison sentence and in total they received four different sentences, it might be, like, three prison sentences, one community, or it might be one prison, two community, but what it's really getting at is it's speaking to the complexity of this group and, I guess, the level of their offending and the things going on for them there.
We also looked at some wellbeing indicators for these people between the ages of 17 and 21 to see what else was happening for them. So you can see that for benefit status, these statutory groups, actually quite a large proportion of all of them received a main benefit at some point but it was particularly high for that crossover group, the YJ only and the CP only, there was not a huge difference there and that crossover group was also more likely to live in social housing for some of that period.
So, for education and employment between the ages of 17 and 21, a very high proportion of those with statutory involvement in either care and protection or youth justice have not been in employment, education or training at least some of the time during early adulthood and yet also a high proportion of them have also been in employment, education and training at least some of the time. If you compare that to the no involvement population at 54% here and 99% here, these just kind of speak a little bit more to instability, that they're not consistently in some form of education or employment and the enrolment in bachelor level tertiary courses is very low for these youth justice groups. Low for care and protection as well, but only 1% of that crossover group was enrolled in university level education.
Again, the mental health and ED and hospitalisation stats are much worse for those youth justice only groups. You'll notice these are a lot higher than some of the earlier mental health stats that I showed you and that's probably just because as they've got older there's been more opportunity to identify issues that need to be addressed so we do expect to see an escalation in those as they get through life, but still staggeringly high.
So that's all I have for you today. But -- so the main things to take away from this, really, are that just because someone goes through youth justice doesn't mean they're going to end up in prison, so there's a -- 24% of that crossover group and 11% of the YJ only group. The crossover group is small, but they have the most complex factors and needs so they're a very important group to look at and help. If you want to see all the research underlying this, because there's a lot more detail and discussion, you can go to our website and it's published on there alongside with a piece from YJ Ops which kind of details some of the programs and things they're doing that speak to some of these results, like supported bail programs and if you have any questions that you don't get to ask today or you just want to have a chat, feel free to email us at firstname.lastname@example.org.