Educational Insights into Children in Care - video transcript
Dr Duncan McCann
Tena Kotou Katoa. Thank you for the introduction there, Kerry. Yes I'm Duncan McCann from the Evidence Center and I'm here today to talk to you about the Educational Insights into Children in Care.
Now essentially what I'm just going to talk about is a few sort of like trends and kind of like general insights that we saw from what, if we look at the data, what does the educational perspective look like if we are just like focusing in on children in care. So let's launch into it, shall we?
So first off I'm just going to have here the IDI Disclaimer. So we are going to be using data here today that came from the IDI. And this is just generally to let you know that there isn't any identifiable information in here, and also these aren't necessary official statistics as a result.
Now a lot of the data is actually sourced from Oranga Tamariki’s Children's Well-Being Model that we built in the IDI. Now essentially this well-being model was developed at the -- for the formation of Oranga Tamariki to provide a measurement backbone. It was really so that Oranga Tamariki could kind of understand what do children's well-being actually really look like, rather than just what our organisation says about them, but what can we see from a kind of wider perspective.
And through consultation with multiple organisations and government agencies over the course of several years, we came up with the sort of conceptual idea of what well-being looked like and that sort of is broadly held in the model under these five well-being domains.
So you've got sort of safety - children safe and feel safe. Security - that's sort of financial resource security. Connectedness that’s do children know where they belong and how they connect to each other in their culture. Wellness - which is broadly a health domain, essentially it's like are children mentally and physically healthy? And then you’ve got development and this is broadly the education domain and that's like do children have the skills and aspirations, skills and knowledge they have to in order to achieve their aspirations.
Now because we needed the sort of broad view of not just like some piece of data here, some piece of data there, to actually really get this full conceptual idea of well-being we had to build the model in the Stats New Zealand's integrated data infrastructure - the IDI.
Now, just for anyone who's maybe not so familiar with what the IDI is, it's basically a giant database that Stats New Zealand holds with almost everyone in New Zealand in it, and they're all anonymised within there, they just have a random ID and you can't actually identify who anyone actually is.
But from the IDI you can see a lot of micro data - you can see the various interactions an individual has had with multiple government organisations and departments over the course of their life going back years and years and years.
So it allows the opportunity to do a large scale modelling of large populations and to really sort of understand what the full kind of life course and interaction of individual looks like. So we use the data in the IDI to feed up into all these various domains, and that creates a whole bunch of indicators. So there's data and there from Oranga Tamariki, there’s data from Justice, data from MSD, health, education and multiple organisations so you get that kind of full picture.
But today we're going to focus in on that development domain – the one that is education because we're only interested in the educational aspects of children in care.
So, first up, what we're gonna do is from the model we pulled out essentially all five to 17 year old children at a particular point in time. Those who are like, you know, eligible basically to be engaged with education, and we looked at one population which was all children who were in the care of Oranga Tamariki currently at that point in time.
Then we took another population - all those who had no care experience whatsoever with Oranga Tamariki to provide a comparison group. And now here we just have a little bit of educate -- a little bit of demographic information and as you can see from the age profile graph they're basically the “In Care Population” and the “No Care History Population” broadly - very similar. They have the same age breakdown so there's not much difference going on there. It’s a bit of a different story when you look at the ethnicity makeup of those two populations.
Now ethnicity as defined in the model is self reported. So what we look at is what is it that you said your ethnicity was. What do you identify with? And so that sources like the census, education, health, things like that. So if you said, and we have four broad categories essentially, if you said you're Māori we consider you to be Māori. If we said you're a Pacific, we consider to be Pacific. However if you said you're Māori and Pacific in one or multiple sources, we have a third category called “Māori and Pacific” because we don't believe we should say which one you identify with most if you've reported it in multiple sources. And then we have New Zealand European and multiple other ethnicities and a broad “other” ethnicity category.
Now if you look at the “No Care History Population” broadly that looks similar to the majority of the New Zealand population - 21 percent Māori, 10 percent Pacific and 4 percent crossover Māori and Pacific.
However, it's quite a different picture when you look at the “In Care Population”. There's a large overrepresentation of Māori in the Care Population – you have 55 percent of kids in care are Māori with a further 8 percent who are Māori and Pacific. So there is a large disparity that exists there in these two populations.
Let's, so we're interested in what does the educational engagement of these two populations actually look like with the “In Care Population” versus the “No Care Experience Population”.
Now the easiest way actually to see what educational engagement might look like in the IDI is actually to look for things that might break that educate, break that engagement that you're having with education. So we're going to focus in on like five sort of disruptive indicators as it were, like we're going to look at, like, how many, what proportion of these populations had stand downs in the past year, what proportion of them were suspended in the last year - like formally removed until a suspension meeting was held. How many had truancy in the last year - that's between what has been defined one in 83 calendar days truant within the school year. High truancy – those who've had more than 84 days in the last school year.
And then, of course, we're looking at alternative education as well. And this is essentially if you're enrolled in education where mainstream schooling is no longer able to actually meet your needs. So if we look at these five indicators and compare these two populations we're looking at, we actually see quite an interesting picture here. We see that essentially the Care Population has a very high representation in all of these educational disruptive indicators. What you can see is like for instance in the last year 9 percent of the children in care had been stood down from school as opposed to only one percent of the No Care Experience Population. Six percent of them had low-to-moderate truancy in the last year's as compared to only two percent in the No Care Population; and four percent had high truancy compared to only one percent.
So you see there is a large disproportion of representation of Children In Care in all these indicators compared to the No Care Experience population.
Now it's important to note though at this point that this graph is not saying that because you are in care therefore you have poor educational engagement. What you need to understand is that when a child is having to be brought into care they've reached, they come from a pretty bad situation. They've essentially had a pretty tumultuous life course that's led them to a kind of crisis state where they've needed to be taken into the care of Oranga Tamariki, now so generally the well-being of children just before they come in care looks very poor in comparison to the rest of the population who has no care experience. So there is a lot of other correlating factors here that that could be contributing to this large disparity that we're seeing.
Now, the thing is we can actually dig a little further down into these indicators what does this look like if we look at how this plays out across age groups. So we're going to focus on two indicators - stand downs in the last year and low to moderate truancy in the last year. And we basically looked at what proportion of five to nine year olds 10 to 13 year olds and 14 to 17 year olds in these populations have displayed these indicators. So if you look at the stand down graphs here, you can see five to nine year olds in care four percent of them had been stood down last year. Now stand-downs are a relatively rare event to the best of times but four percent is a quite a high proportion especially when you compare it to the No Care Experience population where you can see it's practically non-existent.
And then if you go to the next age group 10 to 13 years up, you see it escalates very strongly to 11 percent and then again to 12 percent in the 14 to 17 year old population.
So there seems to be this large escalation with age especially in comparison to the No Care Population, and you see a similar behaviour going on in the low to moderate truancy - again three percent in the five to nine year old group for care and then it jumps to six percent and 11 percent. So there's this escalation with age and it's almost diverging away from this sort of representation you get in the No Care History population.
So, essentially what this is sort of highlighting is like educational, sort of, disengagement is happening for very young children in care, and it's then just escalating further with age. So, there's an opportunity basically to get involved as early as possible - try and find wraparound support and try and stabilise that education at a young age, because it's only going to get worse as they age through the system essentially.
Now we can also break this down by another way. What if we looked at how this plays out across ethnicity? So here we have stand downs last year – low to moderate truancies in the last year by our four ethnicity categories. Now I'd like you to focus on the no stand downs in the last year and the no care history graph here. Now, you can see there is a pretty clear correlation going on with ethnicity here. If you look at the other ethnicity category - less than one percent have had stand-downs then it jumps up to two percent for Pacific, then jumps up again for Māori and Pacific and jumps up to Māori.
So, There is sort of this trend going on with ethnicity here for this indicator. But what's interesting is you don't see that same trend mirrored in the In Care Population. You see, yes, relatively high levels of stand-downs in the last year across these ethnicity groupings, but in fact it’s, they're all at a sort of relatively similar level - there's not as much of a sort of trend going on there.
So ethnicity doesn’t seem to correlate in quite the same way for the Care Population. Interestingly enough on this one the Pacific population has a 4 percent - which is quite low. But this one probably might need to be treated with a little bit of caution because when you talk about Pacific children in care who then have stand-downs you’re starting to talk about quite small numbers, so the sample size is getting a bit small – so there might be a bit of variance in that percentage.
You also see sort of similar sort of non-mirroring going on with ethnicity in the low to moderate truancy. You see that while Māori and Pacific have relatively, like, stay at similar levels in the No Care History population. Those with other ethnicity – very low levels of low to moderate truancy.
Yet once you go over to the In Care Population you see much higher again, but there's a bit of variance at the top but relatively flat around the same level for most of them. So again ethnicity is not quite correlating the same way here.
Now if we talk about educational engagement we always want to talk about stability in schooling as well. That's always a major kind of like aspect of this. So what we wanted to look at is, is we want to look at school changes for these two populations. So if you look at how did any of these, what was the proportion of these populations that had a school change in the last year? You can see that 22 percent of children in care had a school change in last year, as opposed to only 6 percent of those with No Care History. So it's indicative that there is a bit of schooling instability in the last year as it were. But this is just one recent school change - what happens if we looked at multiple school changes? So, let's look at, what is the, how many school changes you've had over the course of your life.
So we looked at how many had three or more school changes over the lifetime and 25 percent - a quarter of children in care - had three or more school change in their lifetime, as opposed to only 3 percent of the No Care History population. Now is a pretty stark wide difference here, and so it's showing that -- and what's even more stark is when you consider the fact that we've actually adjusted this and removed those school changes which are necessary. We removed intermediate to secondary, we removed primary to intermediate - so these are actually a real representation of actual school changes that an individual is going through.
So there does seem to be a bit of schooling instability going on with this population. And, of course, no education argument is complete without looking at achievement. So we looked at basically all 18 to 19 year olds in the model and compared those with care experience versus those with no care experience. And what you can see is, again, quite a stark picture – essentially children in care do not achieve anywhere near to the same level as children not in care. In fact, 45 percent of children with care history achieve no NCEA level qualifications whatsoever, as opposed to only 16 percent of the “those with no care experience”.
Now you can start to unpick the sort of story a little bit more when you actually consider what age children might have left school at. For instance, in the Care History Population 37 percent of them had left school at age 15 or 16, and that's compared to only 10 percent of those with no care experience.
So it's looking like perhaps those in care just weren't necessarily at school long enough to achieve NCEA levels because they are four times more likely to have left school by these ages than the Not In Care Population, which sort of again maybe harkens back to that sort of schooling stability thing we saw on a previous slide.
And it's also very interesting if you actually break down achievement by ethnicity. So what we have here is a graph showing the proportions of these populations who achieve NCEA level two or more by ethnicity. Now if you look at the No Car eHistory graph, you can see there is a trend with ethnicity going on here. Māori have some of the lowest levels of achievement - only about 68 percent of them achieved NCEA Level 2, then that steps up to Māori and Pacific – 70 percent, Pacific 75 percent and other at 81 percent.
Yet when you look at the Care History graph there you see that the differences in ethnicity almost shrink to absolutely nothing. You don't have that sort of same correlation with ethnicity at all there. It seems if you are in care you achieve essentially at the same rate regardless of your ethnicity, which might be indicative that by the time you've reached the sort of, like, situation where you need to be in care, that sort of becomes the kind of overriding factor that determines your educational achievement at that point. Like, the situation around you is at such a state that all kids in care sort of achieve equally as it were.
So it might make a good case that children in care might need that sort of extra support, extra wraparound services, because they're all equally as in need of assistance in this area.
Now, we've looked at children in care, we’ve looked at children without care, but we want to know maybe a little bit more. What happens if you've been in care for quite a long period time? Let's say you’ve spent two or more years altogether in care. Now this could have been a consecutive block - this could have been a couple of blocks over the course of life, we're just looking to see had you been in care for over two years in your life. And what you see, is you see a little bit of a change here. While children who spent two or more years in care still have relatively high levels of these disengagement indicators, it is improved from the general in care population - there only 7 percent of them are being stood down rather than 9 percent. Only 3 percent them have high truancy rather than 4 percent, so that's quite interesting, and it's, if you look into this school stability aspect as well, let's see what do the school changes look like? Again you see an improved picture. You see that school changes in last year is only 13 percent. It's still quite high compared to the No Care Population at only 6 percent, but it's improved the No Care population, so it looks like in recent years children who spent a lot of time in care have had more school stability.
Of course, if you look at the three or more school changes it's still relatively high, but potentially this could've happened further in their past. But there is looking like there is a bit more stability for this population. And when you look at the educational achievement picture of it children spent two years in care still don't achieve anywhere near as well as those who have no care history.
However, they are slightly improved over the In Care Population at a point in time. You see that 42 percent of them, more of them are achieving NCEA levels than were in there In Care Population, and coupled with the fact that you actually see only 32 percent of them have left school by age 15-16, this is again improved over the care population.
So we're starting to see that kind of aspect and maybe they're staying in school for a little bit longer – maybe they have less school changes, so there, it might be a bit of school instability involved here – that would be very interesting story to sort of dig in further and unpick to see if that is perhaps the reason why this is going on.
Now, finally another interesting thing you can do with the model is you can look at sort of like life outcomes for a model population essentially. Because the IDI contains a vast array of historic information for adults going all the way back to their child years you can see what do populations that have certain sort of interactions in their childhood look like in the future essentially.
So here we have basically a graph that showing the proportions that are involved in, like, life outcomes; so, like, participation in education, employment training up to age 20, long term benefit dependency up to age 25, and regular offending up to in your early 20s as well.
And generally what you see is, if you have care experience you have generally poorer outcomes – you have, like, you have less, a lower proportion of them end up participating in education, training and employment. You have a higher proportion of long term dependency and a higher proportion going on to regular offending.
But we're interested in the education aspect of this, what happened is, so we thought what do we look like; what does it look like if we focused on those who achieve NCEA Level 2 versus those who didn't achieve NCEA Level 2 in these two populations. And what you actually see is that, for instance, in the care population if you achieve NCEA Level 2, this dark blue bar, you had a much higher proportion went on to have participation in employment, education and training. A much lower proportion - 26 percent drop down to 13 percent, who ended up on long term benefit dependency in their early 20s and 8 percent of the care population dropped to 5 percent in the regular offending category.
But interestingly enough you see that exact same trend being mirrored in the no care experience population. So achievement of NCEA Level 2 appears to, like, enable better life outcomes. And there could be a number of reasons that are sort of contributing to this, but it would be something very interesting to sort of dig into a little more and find out what is necessarily sort of going on in this aspect here.
So after all that, I just want to throw up a few conclusions for you here about some of the things we talked about today. Essentially we've seen that children in care basically have higher levels of educational disengagement and this tends to escalate with age across the indicators.
And you also have, they have low educational achievement then non care experience children. Ethnicity also doesn't appear to correlate quite as strongly in the care population as in the non care population – both in the indicators and also in the educational achievement. We also know that children who spend, like, longer in care seem to have better educational engagement and achievement, and the fact that this would seem to be coupled with less school changes and a higher proportion leaving school later in their, in later years, maybe suggests that there's a possible stability aspect that needs to be investigated.
And finally we looked at model life outcomes and these were basically showing that these outcomes are better for children who have, like, no care experience as to those who did have care experience. However, this was improved if they managed to achieved NCEA Level 2 regardless of whether they had care experience or not.
So thank you very much for that, and I will be passing over to Karen in a moment and she'll be telling you about those, all those interviews they had with children about their experiences which will be very interesting I think.