6 November 2009
The role of family policies in the promotion of child wellbeing: Lessons from the OECD report Doing Better for Children
Abstract, slides and audio of presentation
Edited transcript
The following audio presentation is brought to you by the Australian Institute of Family Studies as part of our monthly seminar series in which we showcase national and international research related to the family.
The seminars are designed to promote a forum for discussion and debate. They are open to the public and free of charge.
Seminar facilitated & speaker introduced by Dr Matthew Gray.
Dominic Richardson:
Okay, well it's a great pleasure to come and speak to you. I've been looking forward to this for some time. I've had plenty of meetings since I've been in Australia, but I haven't really cut my teeth with the academia. I expect there's one or two people in here, so please feel free to interject at any point. Let me know when I'm making a mistake, saying something stupid or, if you like, something nice, something good.
There were many things that concerned me about coming here and speaking to you. You know, everyone gets a bit nervous and I'll have the usual five minutes of fuzziness before I get into my – see, here we go, that's part of the five minutes.
But then I saw these evaluation forms, and oh dear, I remember tutoring students and worrying so much about the evaluation form. So be kind, be kind and I'll be kind to you if I can.
Not to really wave this around, if you haven't seen it, have a look. We think it's an advance on previous cross-national comparisons for child well-being, and I'll talk a little bit about that, but because I've only got, and I say only got 50 minutes because I tend to talk too much, we'll get straight on.
I'd like to say a little bit about WikiChild at the end. I didn't know what to call myself. I was given the project so I can call myself King of WikiChild or something like that, but coordinator had to do, but I'll tell you a bit about that at the end.
So Doing better for the children covered several areas, including this policy amenable framework for child well-being. We looked at spending in different countries across the OECD, and not just how much but how it was spent and what policies were in place.
We looked at the policies for children from conception through to kindergarten in a chapter. That's the first time that's been done across OECD countries.
We reviewed the evidence on the effect of growing up in a lone parent family on children's well-being outcomes. I'm not going to present that, and intergenerational transmission too, and I'm not going to present that, but I'll take questions on either of those. So I'll do this in order.
We have developed a policy amenable framework for comparing child well-being outcomes across the OECD countries. It's policy amenable because one of the concerns with the UNICEF report, not only was how these things differ or why these things differ, neither of these questions were answered by UNICEF, but there was a response, and this is not my personal opinion, there was a response from governments that some of those indicators weren't their job, and I'm talking about children's relationships, I'm talking about subjective well-being.
As the field grows there's many different ways of focusing your data collection / framework for child well-being. We could do with different frameworks by age, for instance. We know some indicators may well be more important for different periods of childhood. We could do different frameworks for different subpopulations within society in an indigenous group or a migrant group.
This will teach me to speak too long. If it goes off I know I should move slides.
The other things we did is we looked at 30 OECD countries, which is the first time that was done. We looked at housing and environment data. We used HILDA data for that, as well as the European Survey on Income and Living Conditions.
We don't rank overall, and you may have seen - I don't know how well you know the UNICEF report - I know [ARACY] has done something recently using that data, but because Australia wasn't in the final ranking, I don't know exactly how much you know about it, but it put a league table of child well-being together, and this is very, very good for advocacy but not very, very good for policy. For all sorts of reasons it creates perverse incentives. It doesn't tell governments where they ought to be focusing their attention. So we don't do that.
It's more up-to-date, but there are some problems and we're data driven at the moment. We don't have the sorts of survey that would tell us about all the different aspects of child well-being. We don't have the sort of micro-level survey where we could look and suggest how one indicator impacts upon another or what the pathways are between these things, and we can't talk about weighting and we can't talk about - well basically the first point of intervention.
Cross-nationally most child surveys are fit for a purpose: education or health. They concern themselves with well-becoming of children, how they will do in adulthood, not so much how they're doing in childhood, and also to do with child capabilities "Children should only be answering surveys past a certain age". So we're still too adolescent focused. We can't really segregate the index by age or sex.
No one is particularly interested in comparing child protection outcomes across OECD countries, or mental health or neglect, and because of the lack of the sort of micro-level data you need to sort of create robust sort of weighting estimates, we use equal weights, but that's no big surprise because no child well-being index doesn't.
This is where Australia stands in our framework. Housing and environment is where you do well. That talks about the indicators, in that dimension are overcrowding and local environment conditions, and I mean things like pollution, graffiti, noise around houses with children living there.
Risky behaviours is still around the OECD average. We didn't have data on drinking and smoking, and I know you do collect that data, but it's not comparable to the HBSC survey, and I'd be interested in talking about HBSC because it's a useful survey for comparing children's health outcomes across the western countries but there are concerns about methodology. If anyone's interested, we can talk about those.
It does include teenage fertility rates, and as I like to say, for the few countries that only have teenage fertility rates here is that you have two children when you talk about teenage fertility, not one, and so maybe it's appropriate to keep it in.
When you develop a multidimensional child well-being index, you have to have some idea about what your final composite is going to be. You have to have some idea about what you expect the different indicators within that composite to do for each other, and there's a distinction between a composite which is made of both unassociated variables and a composite which is made up of associated variables or internal reliability, something that might be appropriate to use principal components analysis or some clever stuff.
We see child well-being as a causal indicator. It's created by many different - not necessarily interrelated - things. If I was to pick someone out of the audience to help with the evaluation form and was to say something really nasty to them, I'm bringing down their well-being. I'm not necessarily going to bring down all aspects of their well-being. It's not going to have an equally proportionate effect on everything.
So we don't really want to be looking at advanced sort of internal reliability [Cronbach Alpha's ] testing, or that sort of thing, but it's important to remember because we receive criticism when a statistical person comes along and says, well, you haven't done this, and they do it and they get different results. Well, [Hishmartie's] Heshmati's work, it's affected me.
Equal eigths we've talked about. Penalising variation is important. It's a methodological issue for composites. Imagine you have a country which has a health outcome and an education outcome and the health outcome is very good but their education outcome is very bad, and there's another country which has averages on both. I mean which should we value more? Should we value one more because they average out equally to get the same score? These are all things to consider, particularly when our goal is to influence policy, and there's the issue of different numbers of indicators for each dimension.
We select indicators based upon whether or not they're child centred. It's difficult, as you'll know, to come up with consensus between your States on any one indicator, I suppose, or any group. I expect that if we sat here for a day we couldn't agree on what the best set of indicators would be for child well-being, and when you look at 30 different countries with 30 different traditions and histories and cultures, you need to find a common set of goals, and we use the UN convention on rights of the child for that.
Taking another step beyond the UNICEF work, we look at conceptual complementarity. Within every dimension we want to know about efficiency as well as equity, so not only do we look at the mean or the average literacy score, but we also look at the distributions around that average. So a country's system's good at bringing all children up on average and are they good at making sure that they don't leave some children behind.
That's the other issue. Their rights versus development is the well-being, well-becoming argument. We know that children have the longest lives potentially, they have their adulthood to consider, but we also need to make sure that they have good lives as children.
This is the statistical coverage. I hope you can see the difference between the light green and the dark green. The dark green means that we have stage specific data. So we have stage-specific data for late childhood in all of these dimensions and that talks to the - what was I saying? That it's still too adolescent focused.
The light [green] means that we have some indicators - we have an indicator which talks about children from 0 to 17. Governments are still very good at talking about children from 0 to 17. You can imagine, and it's just a fact, from 0 to 17 you develop in 17 years more than you will in any other period of your life. Your needs will change, your capabilities will change, but they're [children] treated as a lump and we deal with that in the report.
You'll notice that education, there isn't any cross-national data in early years, or that's the white. Risk behaviours were, for some obvious reasons, maybe the case. We also wanted to break down by age per year, sex, migrant status, and this is where we managed and where we didn't. Really the main point is that no one segment has ticks across all in dark green. It's just not complete anywhere and that's after we went out and selected these dimensions on the basis of what was available. The red line to the things is the area that Australia has no data comparable cross-national data.
It was very interesting when I was talking about this before that I said surely you have data on bullying. Of course you have data on bullying for Australia, and someone said, well no, we haven't. I mean, do you? It seems to me to be very important if you're concerned about children's lives to ask them whether or not they're safe in school. In fact, you have an obligation, you've signed and ratified, the UN Convention of the Rights of a Child.
Okay, so here's educational well-being. Here's an example. Those are the indicators we use. They're true to convention articles, blah, blah, and I'm sure you know this quite well, but if you haven't it's worth picking up the convention. It's just a few pages, short read. It's very interesting. How we deal with complementarity is addressed in every section. Whenever we talk about indicators we talk about complementarity, and as a way to move forward and to talk about how governments can address these differences, variations along these indicators, we always talk about policy levers. We always tell them how they can make their change.
Education inequality is lower for girls in every OECD country. So the scores for the top decile in Australia for girls are about 1.56 times higher than those in the bottom decile. That's not bad considering it's around 1.62 on average and some countries which talk about, what do they talk about? The strength of their education systems, the abilities of children to educate they're way out of their situation such as the United States have much bigger problems in terms of inequality and that feeds through to things like intergenerational inequality.
Health and safety, they're our indicators. It's interesting that health and safety, health issues are always - we always have more data for health than anything else, and it's partly because of the way health is delivered. It's within an institution and the doctor measures the head or he ticks the box, the mother's okay, the baby weighs this much and so on and so forth. We're given an immunisation. I'd like to say it's to do with keeping the eye on the child but I want you to think a little bit about what it's for. Is it to keep an eye on the costs of the immunisations; is it to keep an eye on the very expensive hospitalisation at birth?
In the child well-being literature there's some discussion about whether policies are fit for survival or for development at this moment. I just want you to think about that, particularly when we think about how to advise governments to change their maternity pay parental leave, and this is a big issue in Australia.
Convention articles, again health is big in the convention, and one of the ways you can look for weighting and how governments want to prioritise different aspects of child well-being is to go to the convention and say what does the convention say. You know, concentrate a lot on education, concentrate a lot on risk or something like that. Well, they're pretty much even except for health. It's the first among equals. It's about - what do they say - they say that health should be provided to the highest attainable standards, whereas they don't talk about standards with the other indicators. They talk about threshold, primary education, adequate standard of living, but highest attainable standard is a different type of requirement.
Comparisons are very important for that, because if you don't know what someone else is doing, you don't know what you can attain. There's no point of comparison.
Again, complementarity is addressed and we have all of these policy levers. It shows I've been talking too long on this slide too. This is good, I shall do this, set my screensaver, move on, something like that.
So here's the indicator example data for infant mortality. Australia's infant mortality rate is below OECD average. The OECD average is brought up a little bit by Mexico and Turkey, but when you compare yourself to the UK and New Zealand and the United States, you're doing okay. In all countries, as you can tell from the trend at the bottom, infant mortality is going down, quicker in OECD Europe, if you like, and the European countries than anywhere else.
Low birth weights happen to have a different trend. Australia's around about the average in terms of low birth weights. Now there's obviously qualitative issues and differences between comparing infant mortality and low birth weight. I mean low birth weight we know, there's literature surrounding the outcomes for children with a low birth weights in terms of their earnings capacity and educational capacity.
It's a concern for OECD countries that low birth weights are increasing and anyone who works in the health area will probably want to tell me to stop making such a big deal because we save babies earlier, and pre-term babies. It's likely this is going to happen, but still whether we save these babies through luck or good judgment, they're still likely to have difficulties in adulthood, and it's important to keep an eye on this population.
Somebody said yesterday in the ISCI conference that we shouldn't be measuring GDP as a measure of society's progress; we should be using low birth rates because prior to the global financial crisis, which I believe hasn't touched Australia like most of the countries, the US low birth rate was around about eight and now it's around about 12. So there's been a big jump. It's interesting.
We move on from outcomes, and there's lots to read there and have a go and send me the email if you don't like it or you like it or you think we should change something, or if you know how I can fill in the gaps for Australia, that would be great.
One of the things we wanted to do was move beyond the simple, this is what it looks like, hold a mere advocacy, off you go governance, do your thing. You know, no help. Solve these problems. We need to provide some guidance as to how it's done. So we review the patterns of public expenditure on children, and we do this by taking the social expenditure which is reported to the OECD and the policies that we know each country has in place and the population figures for children, or where we can enrolment in childcare, and we share out that money. So if you pay 10percent more for children who are slightly older, you give them 10percent more in total but you share it out between all children. It creates profiles, and I'll show you Australia's in a minute.
I have some questions for you because I can't answer them about the Australian profile, and I'd like you to answer them for me before I go because I have to tell when I get back, my masters will want to know what I've learnt and you're going to help me.
Timing matters for child well-being, from the literature that we reviewed and things like when families break down, the times children enter or leave certain forms of schooling, there's effect on them, issues of mobility or their capacity to deal or understand with problems like family breakdown.
Also there's the argument that if you allow gaps or differences to develop in early childhood, the capital which the advantaged children have means that they can use the next dollar more efficiently and gaps increase, so it's about - not talking too much, but it's about intervening early.
I said to the guy on the plane yesterday, and that was my 7th lecture, on the plane to a guy who was very interested in this, and he was a businessman and he has a big company and he employs people, and we talked about maternity pay and he was saying - well, I was saying to him, if you're a businessman, you're interested in cutting your losses, okay. If you want to cut your losses in terms of your tax burden, you want to intervene early because if you wait five years you're going to be paying more to do the same job. He goes no one ever says it like that. I said, well that's why I'm here, so I'm going to say it. Tell your friends.
So testing the Heckman proposition is the other point. Now we all know Heckman's done the [periscope, perischool] analysis, but he also came up with this wonderful parabolic about education intervention and what should be spent when and how future investment would rely on previous investment and become more efficient so you could spend less and early childhood comes first. He had other rules and considerations that are in here.
We thought that was interesting. We wanted to see which governments were spending early. So I told you how we did that.
Also there are issues of costs of children. We know the costs of children differ and some people say, well, older children need pocket money and older children eat more and older children need bigger clothes. Well, younger children change clothes quicker and younger children also have special dietary requirements which may add cost, but the biggest thing that was missing from all of this discussion was parental leisure time, parental opportunity cost. Because you can't go to work and earn an extra $100 a week if you have to care for your child an extra 10 hours a week.
There is the equivalence scale, and the one that's most commonly used and certainly used in European comparisons for child poverty measures gives it an extra weight for older children. You need more money for older children, it says. I know that the use of relative poverty measures has been a discussion in Australia, but among the concerns is the equivalent scales just treat older children as more expensive without consideration of opportunity cost.
Most OECD countries pay more for older children than younger. Austria, Belgium, France, Luxembourg, Mexico, Norway, Slovakia and near neighbours New Zealand all have increments for older children in their main child benefit, universal or means tested. So they're assuming that older children cost more, but Bruce Bradbury has done some work on the costs of a carer, the time costs of caring for children, and you can see under 3s is 26 hours, 3 to 4s is 19, and as you get older a bit more.
In the USA they think it takes more hours to care for their children, but the trend is the same.
Most OECD countries take their family benefit package, their 100percent, chop it into three parts. The smallest part tends to always go to the youngest children; the middle part to the middle children and the biggest share on the oldest children. Now this is interesting, and I'm sure that people will say, well compulsory education is expensive and that's why it's the case. Well, compulsory education may be expensive, but it doesn't stop us asking if compulsory education is efficient or appropriate or part of an iterative process for children's well-being.
It's important because it's the main vehicle for child well-being, which is why it's included for government intervention, but the shocking thing for us, which is why we get to speak to so many people, is when you ask governments why they spend more in terms of their child well-being goals, they don't really know. They don't really have an answer. It's maybe more to do with old goals for child policy which is to free up mum and dad to work or simply to prepare children for adulthood and not so much to prepare them as they develop socially and emotionally through childhood itself.
Australia's is a straightforward pattern. Iceland spend more on middle childhood; Finland spend a bit of money in early childhood. I think it's only Hungary, which isn't on this list, that spends more on early, less in middle and even less in late childhood.
Australian priorities are cash and in-kind transfers. This is 2003 data, but it helps you set yourself in the context. If you take the same amount of social spending and you split up into proportions of that social spending which are transferred in cash or in childcare, with other benefits in kind in education, some countries favour the cash approach – Luxemburg, Germany and so on. Some countries favour the childcare approach - Sweden and the usual suspects. These are the guys that we're all looking at. Other benefits in kind is the Netherlands. Australia is in that top five there, and education, Korea, Switzerland, New Zealand, et cetera. The range shows you the range from the top five, so over a third of one in three dollars in Luxemburg are given in cash to mum and dad. In Australia it's slightly below, but not too far off.
The interesting one, considering your policy innovations at the moment, the childcare tax rebate, childcare benefit, the ideas for harmonising data collection across the States and all of this focus in early intervention, is that in 2003 one in $20 was spent in childcare in Australia whereas one in five was spent in childcare in Sweden. Some of the other patterns are interesting, but if you want to explore them more, ask for the slides at the end.
This is the Australian profile. This is where I can use this. I promise not to…
Question 1: Excuse me, I wonder if you could just really talk about what the benefits in kind might require?
Dominic Richardson:
Yeah, the benefits in kind we're talking about accommodation, child protection benefits, sorts of food supplements, transport supplements and so on.
This is this blue line in Australia. Look, this dark blue line is the - anyway, this is the Australian profile, and this is what you spend in pre-natal years in 2003. Zip, nada, nothing, you know. It goes to 27 because in Germany in 2003 child benefits were paid to people aged 27, to children aged 27, and they also paid them to children, advantaged children aged 27 who were in higher education. So we've got a lot of money. You know, we're going to decide how we'll spend our child benefits. Let's give it to 25-year-old academics.
But these are the sorts of questions. I'd like to have an answer as to why they did. It's to encourage people to go into higher education. Well great, that's great. Some people only have the capacity to go into higher education. It's select out people without the capacity, the background and so on and so forth.
Questions about Australia. Unique things about Australia. Australia doesn't spend very much according to its wealth. This is the hundred per cent of the median working-age family income, okay. So all of these profiles are relative to what the country can afford. Australia peak is transferring about 40percent of that amount in public spending. I must say that the biggest concern with this profile is that we don't pick up some small amount of States childcare here. I'm told that it wouldn't make a great deal of difference.
Australia has this interesting lump here, and this is my big question, a question I promised to ask, which is why in family tax part A do you pay a little bit for children to 12, and then you pay a big bit for children aged 13 to 15, and then you pay a lot less for children aged 15 to 24. Children aged 24, there we are. Why is that? No other country does anything like that. They all go up or down, or stay flat, but no one sort of jigs back and forth like Australia does.
So what's special about your 13 to 15 group that deserves this extra payment? Family tax benefit part B kicks in here where you have an early increment for up to five, and that's what this extra bit of cash here is, but yes, what's this spending about here? I mean, it's not a lot of money, but you could put it there. This you can put it there. In fact, you could take this and add it to this and pay a prenatal benefit. No, you're not doing that. It's in the convention. I should come around and slap all of your wrists.
Comment 1: [Inaudible] is part of the restructuring of payments, recommended increasing payments for that group. You'll also remember at this time that school rates, retention rates are falling. There's a great deal of emphasis on more income support for around people as an incentive to their parents to keep them in school longer. Australia has quite a marked shift in terms of school retention rates. [Unclear] policy change.
Dominic Richardson:
That's good. Now I understand it. It seems to me to be more sensible to do it that way around than promise money when you get there like in Germany. If this is a little boost to ensure people are supported to keep them in school and so on, that's interesting, and it's been evaluated?
Comment 2: [Unclear] which really looks at middle childhood, adolescence, as the time where you will change motivations, attitudes and aspirations around education and life career.
Heckman's recently re-jigged that. That would be my understanding.
Dominic Richardson:
Now I have some reason for this. Australia does it, it's unique. It would be nice to have it evaluated. Then it would inform whether Heckman's re-jigging is any good. We kind of agree with Heckman's second approach, which I'm not fully aware of, but we do say that you should be supporting children here too and at the end, but...
Comment 3: He argues you get positive benefits in terms of cognitive development and you get those middle childhood, early adolescent benefits in terms of retention at school, higher rates of passage from school into higher education. So the model is not as clear as the parabola.
Dominic Richardson:
Yes, I agree. Okay. There is another question? No, I think that's it.
But it is this spending that I didn't understand. I still am not sure what the - I mean, the justification there, but we can talk about that after unless anyone wants to - okay.
Here's France, childcare in France. Now this is the [Maternelle] system that's generated this cost here, which is essentially equivalent to the universal access scheme, but it's from 3 to 5 in France, and they have enrolment rates which are very high. Close to 100 per cent.
You can see the difference here too that there's prenatal spending here and the replacement rate was about 70percent of earned income for mothers before birth. There's also a good deal of - there's a child raising allowance in here, there's an [allocation pour juene enfant (APJE) ] in here, which is part of their home childcare, so mothers can employ someone to work in the home, one-to-one care for the children and it gets paid, and I think the family also gets a little bit of bonus for using that service.
They get up to near a 60 per cent of their median wage, and you can see here in secondary school, you can chop that hump off. They do something that Australia doesn't, and yet to understand the real reasons for this, they increase their spending on children in secondary education by about another 50 per cent. So all of a sudden between 11 and 12 you become 50 per cent more valuable or whatever, but you deserve that extra bit of spending.
Their family benefit's around about here, 20, and there isn't much - in fact, there is no change in the treatment of children by age in their family benefit.
But these profiles don't tell us everything we need to know. They don't tell us about family type changes, they don't tell us of family size. So if there's a payment for lone parent, lone parent's supplement, in your benefit, for instance, or if you pay more for bigger families, as France do, and if they're means tested that's another issue, it's not got private spending in, but we argue that we're talking about being true to the convention and so forth, so in that sense it's a government responsibility so we talking more about public than private, and it doesn't talk about take-up or variation in take-up. Some families may be better at taking up their benefits when children are younger. It also, because it's take-up is a whole, it doesn't tell you exactly what each country is committed to.
We'd be interested in putting those lines in and we are going to do some work on that front, and it assumes temporal stability insofar as it's all for 2003. So policy changes that children, so the treatment of children at 15 may have actually been very different as they go back.
So we deal with some of these differences by looking at using tax benefit policies just as a bit of evidence that over the years single parents and large families are over represented in poverty statistics. You have to use the UK because we didn't include Australia in the tax benefit analysis, so we do the same profiles but we just do it with cash and tax breaks as children age to get at how tax systems and benefit systems treat families of different types as children age, and we look at eight OECD countries but we will be doing this work again, 2007, this year and Australia will be included.
We take the tax, benefits and wages analysis and we add maternity and paternity parental leave and birth grants, and it's linked to other benefits. So previously we didn't have the sorts of supplements that you'd find in your rent assistance in Australia for children. They weren't included in the profile because it was just family and child policies as so defined.
This is an example of comparing the treatment of single parents to two-parent families in the UK. Now if it was just earned income that was affecting these changes, two parents would earn twice as much as one parent. You assume that they're in full earnings, so the line if there was no tax benefit system, would run along two, okay. Now the dark line is the rich families, 150 per cent of average wage, the dotted line is those on the average wage and the dashed line is the poorest.
Now the closer you are to one, the more equal the income is, so the more the tax and benefit system is doing to equalise the income of lone parents to two parent families. So as you can see in the UK because of the work test, on the income support for lone parents, which is around about 13, lone parents could receive income support which made their income more closer to two-parent families during this period of childhood, the first 13 years.
The UK system for most of that period as well is helping the richer families but just slightly. The ratios are lower than two.
The interesting thing is the upward trends in these suggest that the choices by lone parents are going to affect their income. If they return to work, all of a sudden they're much more unequal here. So there is no incentive for the lone parent to return to work all across this period. We assume that they go into low paid work.
So there's questions there for labour market policies. Here, though, is the second thing, and this is in the early years, around birth and up to three in many other countries, this is where it's all going on. This is where all the variation is. This is where incomes go up and down and families need to be supported at lots of different crisis points.
I move on. We looked at conception to kindergarten policies and we looked at them across from prenatal, birth, postnatal and from universal to targeted interventions, and those are some of the things we looked at. Please refer to the report.
We looked at evidence for different types of interventions on child well-being, and this is where I talk about the survival to development confusion. Our policies that we're using today designed for child development, designed for child well-being and they're designed for infant mortality of low birth weight reasons.
There's some question about the statistic, but the prenatal schedule in Australia is higher than the OECD average and we're not finding any good reasons for those additional universal interventions.
We also think that midwifes, there's evidence that midwife intervention is just as good as an obstetrician intervention around for a normal birth. We're not suggesting you shouldn't have obstetricians and so on and so forth, but these are costly things and if they're not doing anything more for your normal birth, maybe you can reinvest the money. It is questionable whether this prenatal care is actually enhancing well-being. We don't find evidence for variation in this type of care. There's no morbidity/mortality differences after one early scan in four check-ups, so that's about five, and even for high risk mothers with birth weight outcomes. I'm using this too much now.
Okay.
Question 2: [Inaudible].
Dominic Richardson:
No, it's the range. It's the range of the prenatal interventions that we looked at. All we're finding is that those with a risk focus designed particularly to look at a smoking and drinking in mothers, or breastfeeding, teaching, and keeping health care notes, those have - we can find some effects, significant improvements for children's outcomes from those, but the highly medicalised interventions we're not finding effects.
Everyone's keeping mothers in hospital for less time around birth and some countries have more paediatricians than others. We're just not finding evidence in either of those cases for additional days for normal births and additional paediatricians.
This is the Finnish Baby Pack. This is a good example of one of the innovations that different OECD countries are doing for children around this time. In Finland when you have a baby and you've been to the hospital, seen the doctor four times, you can get a birth grant. You can either get it in cash or you can get this. There's certain geographical reasons as to why they might provide this as opposed to cash, that not all of these are easily - you can't find them.
The other thing is these are cheaper to provide as it saves the Finish government money, and they're put in a little box which is the size of a bassinet and you put the baby in that to sleep until he's old enough to go in a bigger thing.
But it's interesting innovation. Lots of countries are looking at in kind grants and so on.
So I'll skip through. These are the associations and this is what we know a little bit about the input to outcome relationships. I'm going to talk a bit about, very quickly, I'm going to talk about this, I promise.
Really, this is the stuff that keeps me up at night and I try to innovate with, but it's nowhere near what we can test with the analysis we've done in this, but it's a start. Obviously, we didn't do input to outcome relationships in the report. We're going to be doing more work on refining the policy spending in the profiles and the information we get is very difficult. We're going to refine our outcome measures and so we weren't confident that any associations we've produced would be superbly robust or really informed, but it's good to get thinking now, so I'll look at how age related spending and policy structures relate to some of the outcomes we mentioned in our policy amenable framework.
This is the proportions of the total spent on these different types of services at the top. So countries that spend more of their total child family policy purse on childcare, report lower child poverty rates. I've blued out those that are above a correlation coefficient of .45 or an r-squared of .3.
In those that spend more on benefits in kind generally have higher ranks and quality of school life, although Australia isn't in that, and overcrowding, lower levels of overcrowding.
An interesting thing is that child poverty spending is not associating. In fact, what we don't find there is probably more interesting than what we find. The cash and tax break is not - those countries that spend more in that way are not reporting low poverty rates in a trend.
By stage, it's less promising. Bigger earlier shares are seen in countries with lower teen fertility rates, but that's a huge big gap between the intervention and the outcome, and a cautious approach required. I don't think we'll ever get to that causal pathway there.
And late years and mean literacy scores, countries that spend more in the late years have lower mean literacy scores, so it's partly because the association with early years is reasonably high, it's okay, .4, but because we don't know whether it's causal, we don't know whether that late spending is a reaction to low literacy scores.
So we're going to look now at different kinds of spending at different stages, so this is the early years spending. This is the proportion of earlier spending that goes on in kind types of transfers and we have an association with child poverty there and more spent on in kind spending is proportion of median income the lower the country is reporting its poverty rates. I mean there's a bit of a blob, but I suppose if we're going to take anything from it, it's the fact that only three countries report higher than average spending and higher than average poverty rates.
The chart beside it is showing the cash wasn't doing the same job.
Okay, in middle childhood there's weaker associations, but again cash and child poverty, but cash, child poverty and other benefits in kind is still quite strong, and it maybe hopefully feed into the debate about different types of intervention with different outcomes. Again, the education anomaly appears.
In later childhood there's fewer still. In fact, nothing was above .45. These are the three strongest. It's interesting to note that if you give - there's the cash benefits here that smoking rates are generally higher, and it turns out that in terms of - well, this suggestion that the money can be spent on risk taking, you know, it's giving kids or families money doesn't necessarily give them a guidance about how to spend it.
This is the anomaly. This is the spending middle childhood to mean literacy scores, the education spending to mean literacy scores. Where's Australia? There you are. Above average, but with low spending. So if this doesn't tell us - we don't really want - in fact, this is pretty useless for a policymaker. What it would tell them is that there's some countries here that are achieving very well with low spending and maybe we should go to them for policy examples.
How the spending is done, so this isn't how much but how it's done. Age related, no associations with advanced material well-being or housing environment, though we couldn't split up and down higher and lower spending. Family-size related, those countries that would have an increment for children in larger families have a lower rank - sorry, a higher rank in housing and environment.
The interesting ones are the means testing countries tend to do worse in these rankings and the countries that do provide an advance on maintenance payments for families, child support payments, tend to do better.
And here's our recommendations. I won't go through them all, but we have some for system design, some for resourcing, we talk about front-loading and risk spending and not taking money off of some children just to give to others because we think it's a good idea, but we think some people who are receiving these benefits are not children and maybe we should start there. Spend relatively less on highly medicalised, and I've given you the reason for that, and especially on children past the age of compulsory schooling education. We're concerned about those long child benefits.
Some things to experiment with. Evaluations, we're short on evaluations, and we know there's ethical issues. We shouldn't be sort of sorting children into lone-parent families and children into two-parent families, but we do know the years governments do change their policies and we can look at outcomes for children either side of that policy.
So we're going to do this again, doing better for families. We're going to refine the social expenditure analysis. We're going to develop a child well-being database which not only includes outcomes but context and spending, and we're going to start a review on the systematic bias and cross-national surveys of children in order to improve the data that's collected and give you some support HBSC, et cetera.
I'm sorry I'm talking for so long, but at last you can all relax. It's finished. Thank you.
-- end --
IMPORTANT INFORMATION - PLEASE READ
The transcript is provided for information purposes only and is provided on the basis that all persons accessing the transcript undertake responsibility for assessing the relevance and accuracy of its content. Before using the material contained in the transcript, the permission of the relevant presenter should be obtained.
The Commonwealth of Australia, represented by the Australian Institute of Family Studies (AIFS), is not responsible for, and makes no representations in relation to, the accuracy of this transcript. AIFS does not accept any liability to any person for the content (or the use of such content) included in the transcript. The transcript may include or summarise views, standards or recommendations of third parties. The inclusion of such material is not an endorsement by AIFS of that material; nor does it indicate a commitment by AIFS to any particular course of action.