In the executive summary of the second Intergenerational Report, released today, it says after noting various fiscal pressures that will build over the next 40 years:
It will be important to focus on the efficiency and effectiveness of government spending
I couldn’t agree more. Unfortunately, the government isn’t heeding its own advice, as the Integenerational Report itself shows. In an appendix on spending projections the Report compares spending forecasts made in the first report five years ago with those Treasury makes now. Back then, they thought that Family Tax Benefits A and B would consume 1.3% of GDP in 2006-07. In the second Report they say FTB spending will be 1.6% of GDP in 2006-07. In a trillion dollar economy – as various government Ministers for some reason keep telling us – that 0.3% is a lot of money. In real per person terms, it’s gone from $613 per person to $790 per person, or about a 29% increase.
Now what do we have for all this money? According to the Report:
Higher fertility and higher levels of migration slow the rate of overall ageing of the population. Policy initiatives that have supported this include expansion of the Maternity Payment and other policies to support families, including large increases in the maximum rate of the Family Tax Benefit.
It’s true that the fertility rate has edged up, but in absolute terms births are only modestly higher, about 4% more, in the latest statistics (2005) than they were in 2002. As Ross Guest points out in an article on the baby bonus in the new issue of Policy these policies tend to be inefficient, because they mostly reward people for doing what they were going to do anyway. Most long-term couples will have at least one child regardless of incentive schemes. At this stage, it is not clear that these family schemes have had any long-term impact on fertility. We could just be seeing the delayed fertility of many women who had left starting their families rather late. As spending programs, family payments are unlikely to be efficient or effective.
For a Report concerned about fiscal sustainability, the amounts involved are not trivial. Though the second Report forecasts a decline in FTB spending per person in future years, even a decade from now the prediction is that the FTBs will cost more than the PBS, a major driver of the coming fiscal pressure. Yet as I argued in my big government conservatism article this was money the government chose to spend, not money it spent for compelling policy or political reasons. But now that it is being spent it will be very hard to cut, because millions of people get a share of it and many of them live in the seats that decide elections.
35 thoughts on “The fiscal burden of Family Tax Benefits”
You’ll still vote for them Andrew, won’t you” Somebody has to.
Never miss a chance to bash the old FTB Andrew. 😉
I’m interested – what are they projecting for the 50% CGT discount? It has blown out from $2 billion in 2003 to $5 billion in 2007. Either swapping it for genuine EMTR reductions at low-to-middle incomes, or just getting rid of it and saving the revenue would surely do more good in the long run.
Completely off-topic – I’ve been trying to find out what the median gross full-time wage is, and having some difficulty. Do you (or any of your estimable contributors) know where I should be looking? Any pointers would be good.
“Policy initiatives that have supported this include expansion of the Maternity Payment and other policies to support families, including large increases in the maximum rate of the Family Tax Benefit.”
This is very easy to claim, but it is no doubt harder to support with evidence! I read or heard somewhere (on the radio?) that the current increase in the rate of childbirth might be moreso due to baby-boomers kids having their own kids rather than anything else. I don’t know whether this is true or not.
Leopold – ABS Employee Earnings and Hours provides that statistic.
I’m surprised that family benefits cost so much – I would have thought the conservative argument would be that families kind of spontaneously form and create security in the face of stormy market forces – by virtue of e.g. two potential incomes – without the need for payouts.
Hmmm … according to Lindsay Tanner, government spending was lower as a proportion of GDP under Hawke than it is now. Perhaps someone can confirm, deny or explain this to help make my decision on who to vote for a little easier. 😛
Also, as a relative newcomer to an interest in politics: is Peter Costello more small-government than John Howard? What about Malcolm Turnbull and other frontbenchers?
Leon – This is the Budget time series, though it is preferable to use ABS statistics.
I’d have some hope that Turnbull would be more restrained in spending than Howard or Costello.
Andrew, nice post. I can’t help thinking that the $4100 baby bonus cheque that the government sent me recently could’ve done more social good if it were directed towards a teacher merit pay bonus.
Incidentally, the comments thread suggests that you would do well setting up a pay-per-stat statistic-finding service. You’re very speedy with it.
Thanks a lot.
And as AL says – you could charge.
Ross Guest also says first and foremost that there is no evidence we need to encourage women to have more children! That’s the most galling thing about all of this pork being dished out.
The broader issue is the counterfactual – what could the government have done with the money. Maybe it could have reformed the tax system had it known at the outset what a windfall it would receive. Would Labor have done better? Labor under Crean and Beazley opposed virtually all reform – GST, Telstra sale, IR. Now Rudd has come out with the idea of picking winners in telco infrastructure. I’m not confident that released from the shackles of the recession and concerns about the current account deficit, Labor would have spent our money more wisely or nuture less of an entitlement culture.
Thanks Andrew – and thanks for a great blog in general. I’m sure my first vote ever will have been influenced by your posts!
By the way, speaking of Malcolm Turnbull: perhaps it’s common knowledge, but I didn’t know he defended the author of Spycatcher against the British government!
Yes, that’s one of the fights that he has taken on. I remember it being said once, he took on Thatch over Spycatcher, he took on Kerry Packer over something or other (I forget what), and he took on John Howard over the republic. Since then, he has taken on the Treasurer over Australia’s tax system (remember how early last year he funded an Access Economics report into our system?)
He’s not afraid of a fight.
I’ve been meaning to make a comment on the “regressiveness” of Family Tax Benefits for some time, but it’s been a while since you have written about it.
Using the standard approach to analysing the impact of government spending on income distribution, family benefits are generally progressive not regressive.
This is because when we analyse income distribution data to measure household wellbeing, it is customary to adjust income by family or household size using what is called an “equivalence scale”. For example, it is commonly assumed that a single person household needs between 60% and 70% of what a couple without children needs in order to have the same standard of living, and that each child needs between 15% and 20% of what a couple need in order to meet their own needs. The implication of using this sort of scale is that a single person household with a disposable income of $60,000 is as well-off as a couple without children and a disposable income of $100,000 and is as well off as a couple with two children and a disposable income of say $130,000.
The further implication of this would be that a transfer from a single person with a disposable income over $60,000 to a couple with two children and an income less than $130,000 reduces inequality, that is the transfer is progressive.
The idea underlying equivalence scales is fairly obvious – each person added to a household has additional needs, but there are economies of scale in consumption (hence a single person needs 60% to 70%, not 50% of what a couple needs), and children have lower needs than adults.
There is an extensive if old review of this literature by me at http://www.facs.gov.au/internet/facsinternet.nsf/VIA/prpspdf/$File/PolicyResearchPaperNo27.pdf
More recently, a lot of work on this has been done by Bruce Bradbury – see in particular http://www.sprc.unsw.edu.au/dp/DP132.pdf but there are other papers on his website.
A number of other conclusions arise in this literature – older children have higher direct costs than younger children, but there are much higher indirect costs in terms of foregone earnings for families with very young children, lone parents may need something more than just a single person and a child, and there are additional costs of working. A further finding is that the costs of children decline somewhat with family income.
It may be possible to completely reject the idea of using equivalence scales to adjust family incomes, but as far as I am aware no serious academic study of household income distribution ever has. Look at the Luxembourg Income Study (LIS) website and you will see that every paper on household or family inequality or poverty uses an equivalence scale of one sort or another.
The ABS also use equivalence scales in their analysis of household income distribution see: http://www.abs.gov.au/ausstats/abs@.nsf/Latestproducts/6523.0Appendix32003-04?opendocument&tabname=Notes&prodno=6523.0&issue=2003-04&num=&view=
Equivalence scales are unnecessary, however, when looking at individual incomes (for example, Andrew Leigh’s analysis of income tax data).
In fact the use of equivalence scales has a stronger impact on measured income inequality than the income tax system. If you look at Table A.5 in the ABS source, you will see that the ratio of gross incomes between the 90th and 10th percentiles is around 8.5 to one, the ratio of disposable incomes is around 6.6 to one, but adjusting with an equivalence scale reduces this disparity to 4 to 1. So if one completely rejects the idea of equivalence scales one is also required to accept that income inequality in Australia is more than 50% higher than the ABS measure.
Now, accepting the idea that larger households need more than smaller households to be equally well-off does not mean that you have to accept the current family benefit system in Australia in all its expensive glory. As you point out there are various objections to maternity grants along the lines that they may not be very effective. And the system may have various undesirable effects on incentives. There is also an immense debate about how you estimate equivalence scales, and whether the common ways of estimating them are meaningful. So there are lots of specifics to argue about.
However, one also shouldn’t simply assume that family benefits are regressive. The system would be regressive – or part of it would be regressive – if it “over-compensated” families with children for the cost of children. As mentioned above, one finding of the literature is that the relative costs of children decline as family income increases. This in fact tends to support the current system rather than having a system of tax deductions which would provide constant relative support to families irrespective of their incomes.
What is unusual (in a historic sense, but not in a comparative sense) in what has happened in Australia over the past few years is that increases in family benefits have been made at the same time as tax cuts. However, if the government makes tax cuts and doesn’t increase family payments at the same time, this means that the relative position of people with children slips compared to those without children, so families could equally argue that previous ways of providing tax cuts were regressive.
I think that what you need to do is to establish that the current system over-compensates families with children – it is possible that it does, but no one here or on your previous posts has proved this. The way to do this would be to model the distributional impact of the system of family benefits using different assumption about the costs of children (easier said than done).
Peter – Thanks for that detail and background. My disagreement with you is more philosophical than technical -I accept the equivalence scales analysis. I’m not convinced that we should redistribute income simply to make the overall income distribution more equal (which the FTBs have done overall, though both do also provide handouts to families I would regard as affluent). I do not however oppose other welfare measures where there are other needs, such as poverty alleviation or possibly even encouraging people to have more children, if the birth rate is too low. Such measures would make the income distribution more equal, though that is not their purpose. In this post, I was pointing out that the Integenerational Report was linking family payments to increased fertility, and disputing their effectiveness and efficiency in achieving that goal.
Peter, I agree with your point that an assessment of poverty needs to take account of need and not just income. But this means that if one accepts Guest’s argument that there is no (economic) rationale for rewarding people for having children, having children is a self-inflicted creator of poverty. So the FTB means we compensate people ex post for making choices that impoverish them. At the same time, the psychic benefits of children – which flow largely to their parents and immediate family – are untaxed. Perhaps we should have similar tax benefits for people who are impoverished because they choose to travel, change jobs frequently or buy plasma TVs?
Universal Declaration of Human Rights Article 16.
(1) Men and women of full age, without any limitation due to race, nationality or religion, have the right to marry and to found a family.
Nothing in there about plasma TVs ……
Thanks Russell, although I checked the UDHR carefully and nowhere did I see the right to take a dump when I feel like it..
Seriously, no one is disputing whether we have the right to have children or travel or buy plasmas. The issue is whether we should shower benefits on people who make certain choices.
But you’re equating the choices – it was important enough in the UDHR to mention founding a family, but not to mention a right to go shopping.
I’ve said this before, but it may need repeating. While it is possible to analyse the choice to have children in a manner similar to the choice to purchase consumer durables, children are not televisions. Children are people. The fact that children increase the welfare of their parents is relevant to tax and transfer policies, but it is not a reason for not being concerned about the material and personal wellbeing of children in their own right.
Parents do choose to have children, but the assumption underlying family benefit systems is that by providing benefits to parents, we are mainly benefiting their children – who had no part in whatever choices were made.
I suppose the broader point that I’m making is that it is certainly possible to criticise the family payments system from a range of perspectives (see the work of Patricia Apps, for example), but I think that what appears to be the assumption – more in earlier posts than this one – that the FTB has virtually no rationale at all – doesn’t really advance the debate.
Peter, I take your point that children are people and hence their welfare is important, whereas it is meaningless to talk about the welfare of TVs. My point was just that having children is but one way to impoverish oneself. The tax and transfer system is blind to whether I spend up big on consumer durables (or disposables), accumulate big debts, have to pay lots of interest and then have a low net income to live on. Yet I, as an adult, am a person too! Shouldn’t a payment from someone on identical gross income to me but who does not have to pay lots of interest be regarded as progressive? Well, we don’t have such payments because they would set up very poor incentives. We just accept that people with large consumer debts will be worse off than people with low or no consumer debts, other things being equal. It’s true that children don’t choose to be born. But if children are born, they cannot be regarded as being worse off than if they weren’t born. So why can’t we be comfortable with the notion that other things being equal, families with children will be worse off than adults without children, just as we are comfortable with the notion that people with large consumer debts will be worse off than people with no consumer debts? And hence that payments from childless adults to families are just as odd a concept as payments from savers to spenders? Sorry, I really don’t mean to be obtuse.
I think you need to break this down into 2 further categories. 1) Payments I make to you with children out of self interest. (i.e., you are poor, and children will turn into trash because they don’t have enough resources); and 2) payments I make to you not because you are poor and your children will become trash, but because I want to reward you for your good breeding behavior. I agree with (1), but I don’t agree with (2). I think (2) relies on the belief that we need Australian born children to solve demographic concerns; or we need Australian born children for nationalistic reason. In addition, as the original article suggests, it might be quite difficult to affect people’s breeding behavior (although you can probably do it if you are willing to risk going broke doing it — like France).
“But if children are born, they cannot be regarded as being worse off than if they weren’t born.”
This may be true, but why is it relevant? Being born creates people, and those people have needs. The question is whether meeting those needs is solely the responsibility of their parents or whether society has an interest in trying to achieve a certain level of outcomes.
I think that you (and Andrew and many other people) are looking at family payments as if they are solely a benefit to the adults who choose to have children. But the relevant comparison of wellbeing is to take account of all the individuals in a household, including the children. One also needs to take account of who benefits over multiple generations.
As I pointed out some time ago, every adult was once a child and has already benefited from some of these subsidies (perhaps not as generous as the current ones); the vast majority – 85 to 90% of adults will have children in their lifetimes if current fertility patterns persist; and the people who never have children will also benefit from other people’s children in an extremely wide range of ways (as other people’s children grow up to fund pensions, health care roads, police and to produce all the goods and services that the population want to consume).
Now you could argue that other people’s needs are irrelevant for policy purposes, but this does not mean that those needs do not exist.
You could also argue that the child’s needs are “outweighed” by the enjoyment that parents derive from children – but this is a type of empirical argument that needs to be justified by some sort of quantification of the benefits to parents compared to the costs of having children. The paper that I referred to in my original post by Bruce Bradbury discusses this.
This in itself does not justify government intervention, but my argument is that to simply make welfare comparisons across households as if only the adults mattered is wrong (as Russell also points out).
I also think that the analogy with people taking on debts doesn’t work at a number of levels.
For example, there are extensive direct and indirect subsidies for various forms of saving that reduce current consumption, including superannuation and home purchase. Now one might argue that these interventions are not justified, but they do exist.
Peter – If your argument is correct, how do we decide the ‘right’ amount to transfer to families?
Now that is the $64,000 question!
The equivalence scale literature and related analysis of the indirect costs of children in terms of foregone earnings and time use data provides estimates of direct and indirect costs. Different methods and different data produce different results, but I think it is possible to narrow it down to a range of estimates.
But then society has to decide what share of costs is a purely private expense and what share it is believed should be shared. Two principles could be that nobody receives more than the cost of raising children, and that indeed all parents should make a significant contribution to their own children, or otherwise we would be socialising the cost with probably adverse implications for incentives.
However, one might also argue that more of the costs could be shared for low income families than for high income families (which is what happens in Australia).
The first stage is a complex technical exercise, but still involving lots of value judgements. The second stage – how should costs be shared – is purely political. One might argue that zero sharing is appropriate, and there are respectable arguments for doing so. (like the link in your original post)
One could leave it up to Parliament but that might not satisfy many people. One alternative is to have some sort of review as with the Review of the Child Support formula and CSA.
This has links to various research papers, and it also involved extensive consultation, but I imagine that consultation would have to be even wider for the whole system of family assistance.
It is also possible of course that one might conclude that different parts of the system are justified to differing extents.
“Now that is the $64,000 question!”
It’s the $64,000 answer I am worried about:)
Yes, most estimates are that parents spend considerable larger amount of money on children than $64,000 – but not per year!
In a nutshell: Children are not televisions, they are spillovers. Government never subsidises televisions but they do subsidise spillovers. So we move from consumer durables to mythology. It’s very unsatisfying.
The system of family payments has a range of objectives.
An early objective was related to the concept of paying tax according to ability to pay, with the idea that people who were supporting children had less ability to pay tax than people without children. Over time, these tax deductions became tax rebates and then cash payments, largely because successive governments wanted to target more of the tax relief to lower income families and to make tax benefits refundable.
In addition, governments were concerned about poverty among low income families with children, and provided benefits to children of people who were unemployed or disabled or sole parents.
Then in the 1980s governments became concerned about work disincentives because very low paid working families could have lower disposable incomes than those on benefits, so they increased benefits for low income working families with children. Under the Accord, workers also traded-off wage increases for these higher family payments.
As a result, the Australian system is now among the most targeted in the OECD. For families with very low incomes, Australian assistance is about the third highest in the OECD (adding cash payments to tax reductions and expressed as a percentage of the average wage in each country). For higher income families over 150% of the average wage, Australian benefits are about one-third of the OECD average.
One could move back to providing support to familiesthrough the tax system; this has been proposed by CIS for example, but if you wanted to keep the targeting to low income families, then a large part of the system would look very much like it is now.
However, as I read the CIS proposal, it actually involves larger tax reductions for high income families than are now currently available. To me this seems inconsistent with the ideas underlying the original post.
The CIS has, if I recall correctly, published work supporting tax benefits for families. But while I work for the CIS this is not a CIS blog and I don’t endorse everything the CIS publishes, not even everything I decide to publish in Policy.
“85 to 90% of adults will have children in their lifetimes if current fertility patterns persist”…”64K”
Getting away from the really poor category, surely this figure supports Andrew’s worry that $64K is too high — as it simply is going to be an everyone subsidy. Based on median wages, this number is going to catch the majority of single worker families and a fair chunk of 1 full time/1 part time working families as well. Why have a subsidy the majority of the population gets and ends paying for? I also don’t think the OECD figures are a good guide because I can’t see how many of the countries at hand are going to afford to maintain their subsidies, particularily with the greater levels of people movement across countries. Its hardly even worth working in some countries in Europe if you are childless, and this isn’t a good way to stop emmigration of a fair number of the most motivated people in your workforce (let alone get motivated people back — something that countries like Australia rely on).
Peter, I agree with you on several things:
1. “the relevant comparison of wellbeing is to take account of all the individuals in a household, including the children. ” and “to simply make welfare comparisons across households as if only the adults mattered is wrong”
Agreed – children are people.
2. I also agree with your second principle that: “all parents should make a significant contribution to their own children, or otherwise we would be socialising the cost with probably adverse implications for incentives.”
Where I disagree is on:
1. Who benefits from FTB and what this implies.
(a) I don’t have data to dispute that 85-90% of people will have children in their lifetimes. But if true, surely that means that (i) giving even a modest benefit to families *really* screws childless adults; and (ii) an awful lot of (inefficient) churning is going on.
(b) Your assertion that childless people benefit substantially in net terms from other people’s children. I’m not convinced this is true – I suspect there is a significant demographic dividend to my generation (X) if we don’t have kids because we will not have to bear the cost of raising and educating them for 25 years before they become productive. (This is not to say that having been born, educating kids is a bad idea.) And what’s the counterfactual? I suspect I would get a better return from my money investing in emerging markets over the long term and importing domestic labour if and when I need it down the track. Because that’s all other people’s kids are to me – future (expensive) labour and a nuisance in the meantime.
2. “I also think that the analogy with people taking on debts doesn’t work at a number of levels.
For example, there are extensive direct and indirect subsidies for various forms of saving that reduce current consumption, including superannuation and home purchase. Now one might argue that these interventions are not justified, but they do exist.”
Isn’t this a non sequitur? What do superannuation tax benefits have to do with consumer debts or FTB? They are completely unrelated. You didn’t actually explain why the analogy does not work. My point was that people can become impoverished for a number of reasons, not just having children. And if you cared only about material wellbeing outcomes, you would give people money regardless of how they got themselves impoverished. But we don’t do this because we think people who incur consumer debts should be responsible for the consequences of their actions. Why not the same for those adults who have children? This point holds even if the money is used directly for the children: Presumably parents care about their children and hence derive satisfaction from them being provided for. Yes, children are people but they are the creation, and I believe, the responsibility of their parents. You seem to agree with this view to an extent (see above).
To my mind, the key rationale for differentiating between families and debt-fueled consumers is a moral one – having children is regarded as more worthy than incurring consumer debt. I disagree that it is, given that people have children for their own gratification, not for the good of the country.
the government increased FTBs, in per person terms, 29% and left multiple opportunities to be socially responsible and receive additional tax benefit from art:
In other words, the government does quite a bit, it’s just the knowledge about this “doing”is with masses…