Income inequality in the noughties – how far would you go to fix it?

In the recent mega blog discussion kicked off by Don Arthur, I ventured the opinion that “the truly remarkable thing is that the Gini coefficient of equivalised disposable income has only increased from 0.28 to 0.31 in the last 30 years of so.”  Given the underwhelming response from other commenters to those words of wisdom I thought I might expand a bit on how I came to that conclusion.

Lets say for starters that we are back in the 1950s.  We have five working to middle class families, with varying numbers of kids.  Lets just suppose that all of the husbands work full-time in similar jobs in a local factory, or bank, or insurance office.  All of the wives are ‘homemakers’ (dontcha love that word?), most having retired from work when they got married.  In that world, equivalised disposable income is mainly affected by the number of kids in the family, but overall inequality in disposable income (the Gini coefficient) is relatively low.

Fast forward to today.  All of the Dads still earn similar incomes, but probably in a wider range of jobs.  In one of the families, Mum went off to university and got herself a degree and a well-paid full-time job.  In another, Mum has a permanent part-time job, say as a teller in the local bank.  The Mum in the third family works as a casual in a local supermarket, because she likes to be available when the kids come home and to take time off during the school holidays.  The fourth Mum still believes that a mother’s role is at home as full-time homemaker and mother.  And the fifth family has split up when the parents decided they didn’t love each other any more. That Mum has moved with the kids into public housing in a nearby suburb – she has a bit of a part-time job during school hours, but is mostly reliant on income support.  Guess what’s happened to the Gini coefficient in this world, compared with the first?

Well, I’ll admit that that is a highly stylised account of some of the changes that have happened in Australia in the last 50 years and that there have been a lot of other changes overlaid on the top of these.  But it’s pretty obvious isn’t it that even if the distribution of wages hadn’t widened, household income would be a lot more unevenly distributed these days than it used to be.  And the other relevant point for me is that the increased inequality has come from people making free choices about their education, work and living arrangements.

So my question for those of you who believe that Australia’s level of income inequality is too high now is this – what would you do to reduce income inequality among this particular set of families? 

This entry was posted in Economics and public policy, Gender, Society. Bookmark the permalink.
Subscribe
Notify of
guest
44 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
SJ
SJ
13 years ago

…what would you do to reduce income inequality among this particular set of families?

Nothing, probably. Why would income inequality within the middle class be a problem that needed addressing? FFS.

NPOV
NPOV
13 years ago

It’s not clear to me that the sort of changes BG is talking about are likely to have a huge impact on the overall GINI co-efficient, given the way it is calculated. SJ’s right – you’re basically talking about incomes within the middle classes become more varied.
As I understand it, the big reason for the change in Australia’s GINI co-efficient has been due to changes at the very top end of the scale – those in the top 5% or even 1%. Income growth in the top 1% for instance has far outstripped that in any other bracket. There is some cause to be concerned about this as a trend, given that the lot of the bottom 1% hasn’t improved terribly.

conrad
conrad
13 years ago

NPOV,

I just don’t see why people should care about the top 1%. As far as I can see, this is basically just jealousy, since the lives of the top 1% have almost no effect on the bottom 1%, or almost anyone else for that matter (except that more taxes can be collected from them). Alternatively, I do see why people want to do something about the bottom 1% (in fact, probably a much greater percentage), but their problems are basically independent of the other group.

NPOV
NPOV
13 years ago

conrad, I only care in so much as

a) If that top 1% becomes so separated from the rest of the country, they risk become a powerful elite answerable to no-one else, as is the case in many third-world countries.

b) It’s the existence of the amount of wealth at the top 1% or 5% level that makes the conditions the bottom 1% live in so shameful. We clearly *can* afford to do a far better job of looking after the bottom 1% than we do. Taking as little as an extra 1% income from the top 1% and giving it to the bottom 1% would make a huge difference to that bottom 1%, and the top 1% would barely notice.
Of course, I’m not really proposing that as a solution; I’m just trying to explain why you can’t ignore the top 1% when worrying about the bottom 1%.

NPOV
NPOV
13 years ago

BG, no I didn’t say it doesn’t have an impact, but I doubt it’s responsible for more than perhaps a third of the change in Australia’s GINI co-efficient over the last 30 years. That’s purely based on what I’ve read elsewhere – that most of Australia’s recent growing income discrepancies have occurred at the top end.

And no, the problems of the bottom 1% may not be “solved by simply redistributing money”, but if done sensibly, it seems to me to be the fastest way of improving the living conditions of our least fortunate citizens (and, accepted, an inevitable percentage that exist in that state as much out of laziness than misfortune).

NPOV
NPOV
13 years ago

BG, bottom 20% is quite different from bottom 1%.
I agree that the bottom quintile as a whole has probably received as much government support as can realistically be expected.

Anthony
Anthony
13 years ago

Nice post, BG. but as you point out at the start, the gini co-efficient has ‘only’ risen .03 points. So if you’re suggesting government and/or other trends have worked to contain what would otherwise be a substantial growth in inequaility of disposable incomes, then the answer to your question on how to even further contain the growth (ie. reduce the coefficient .03 points) would probably be ‘more of the same’.

I’m presuming the containment has come about due to a number of factors that have changed in the past 40 years: increased government payments to families with kids; increased government payments to the single mum in the public housing estate (and allowing her to package this with a considerable amount of earned income); the couples having fewer kids to begin with (probably because they start breeding later). What else do you reckon has been going on? And I’m always unsure whether having more dual earner households is likely to increase or decrease the gini co-efficient, especially in the face of rising wage inequality (especially amongs the husbands in your scenario).

Thinking in old ways
Thinking in old ways
13 years ago

BG – my gut feeling is that the gini is most probably understating more recent changes because of the changes ABS have made which have improved data quality – but that doesn’t really change the essential question you have asked – which is what changes should we be concerned about and which ones we should not.

I am really surprised by SJ and NPOV who consider the lone mother living in public housing with little private income as just being part of the middle class.

The debate about the bottom 1% is interesting – and it is worthwhile thinking about who is at the bottom. In most cases the bottom 1% in Australia are those who are ‘outside the system’ – a mix of those with significant mental health, personality or substance abuse problems that mean they are unable to either claim welfare – or maintain their claim. They are most probably not picked up in the data either.

Above the bottom 1% there is most probably some 10- 15% or so who are welfare dependent – all of whom are getting much the same level of income (with some variations – those on Newstart at the bottom and most probably lone parents nearer the top).

This is not quite the same story that the data tells – but given it is self reported income taken on face value by the ABS and includes small business and other investment losses etc, that is to be expected. There are also households showing up with little if any income because they are being supported by others – middle and upper class students being supported by parents living away from home, recently migrated aged parents of migrants who are being supported by their children, etc which show up as having no income.

This suggests then that the questions about the real bottom are not about income distribution as such – but rather how society treats those who really have problems living in it. Then you have the big lump on income support – all in much the same boat – although there have been some changes within the group – lone parents have gained lots over the last couple of decades, aged pensioners have got somewhat better off – but not as quickly as working people – and as for Newstart while most have benefited from the economy by getting off – for those left on the story is not quite as good.

NPOV
NPOV
13 years ago

Well, no, obviously BG’s last example of the mother forced to move into public housing and rely on income support is not “just…part of the middle class”. But if there truly are a large number of people who have gone from a comfortable middle class household to a poor, struggling one purely because of family break-ups, then it is very much something I think we should be concerned about.

NPOV
NPOV
13 years ago

BG, yes I was tempted to bring up the moral hazard argument, but not to put too fine a point on it, I think it’s pretty bogus. I thought we’d got past the idea that unhappy families should stay together purely for financial reasons. Further, unless you can demonstrate that providing additional support for single parents is radically increasing the rate family break-up, then the consequences of providing insufficient support are surely far worse.

FWIW, I think there at least one very strong reason for providing extra income support to single parents: they are doing an essential job raising children, without which our society and economy would rapidly collapse. The other way to view it is that the support is not going to the parents, but to the children, then entrusted to the parents to spend appropriately.

Anthony
Anthony
13 years ago

“For example, should taxpayers (via the government) fully compensate people for the adverse financial consequences of having decided to form two households from one by splitting up? ”

If two grown adults without kids decide to split up and suffer adverse financial consequences, then the government doesn’t compensate them and I rarely, if ever, hear anyone suggesting it should.

If on the other hand a couple with kids decide to split up, the government will step in and provide some compensation with the rationale of protecting the kids from the adverse financial consequences not of their decision, but of their parents’ decision. This seems fair enough. (Most policy initiatives to improve sole parent payments in the past 30 years were based on the idea that no child should live in poverty etc etc, in the context of a society that, since 1975, had decided that adults should have the freedom to split up)

“How much would you have to increase income support payments by to get it back to 0.28 ”

It is my melancholy duty to confess I have absolutely no idea. Can someone help out here? Peter Whiteford? Anyone? Anyone?

NPOV
NPOV
13 years ago

BG, as we know, even James Packer was eligible (and probably claimed!) the first home buyer’s grant, baby bonus, and various other transfers. Increased means testing for existing transfers, and using the savings to boost payments towards those at the bottom end of the scale would surely help reduce Australia’s GINI co-efficient, without increasing the tax take at all.

NPOV
NPOV
13 years ago

BG, I don’t see it as helpful to think about whether taxpayers have any sort of “duty” to pay for anything. As I see it, the point of tax & transfer systems is to produce a society that everyone can be happy with and proud of. If it can be done by taking less tax, great, but if it really does require a larger pool of funds to achieve, and there’s no real room in the federal budget for alternative savings, then the solution logically has to be increasing the tax take. However, given the tax take has increased from ~24% to over 30% of GDP in the last 11 years with no obvious improvement on the equality stakes, I doubt that’s actually needed.

Peter Whiteford
Peter Whiteford
13 years ago

Since my name has been mentioned Ill put in my two bobs worth.

Like Anthony I initially thought that I wouldnt have the slightest idea about how much you would need to spend to get the Gini coefficient down to 0.28, but then I realised that I do have some numbers that might help. This is based on the 2003-04 SIH, so it will have changed a bit, but not much.

On OECD calculations the Australian welfare state reduced the Gini coefficient from market to disposable income from 0.458 to 0.301 (or multiplying by 100) by 15.7 percentage points. About 2/3rds of this was through transfers and the reminder through taxes. Transfers amounted to 14.3% of household disposable income, so it can be calculated that for each one percentage point of household income spent currently on cash benefits, the Gini coefficient is reduced by 0.68 percentage points, while for each one per cent of household taxes the Gini is reduced by 0.19 percentage points.

(As an aside the Australian system is the most efficient in this sense in the OECD, but not the most effective, since many other countries spend more than Australia, but at the other extreme is Austria where each one percentage point of transfers reduces the Gini by 0.14 percentage points or by about one-fifth the Australian level.)

This implies if you believe all these numbers that if you didnt change the distribution of benefits (and Australia already has the most progressive distribution of cash benefits in the OECD by a wide margin), that if you increased transfers all around and the taxes to pay for them by roughly 3 percentage points of household income, that the Gini coefficient for disposable income would be reduced from 0.301 to 0.28.

Initially this might not sound like a lot but this would mean raising transfers from 14 to 17 per cent of household income (or by a bit more than 20% of their current level) and taxes from 23 to 26% of household income (or by a bit more than 12% of their current level).

This of course disregards any behavioural or political responses!

So yes, it is possible to come up with spuriously precise answers to difficult questions!

I have some other comments, but will post them separately.

NPOV
NPOV
13 years ago

No disagreement there. I really doubt that there’s any need to increase any current income tax rates in order to address Australia’s more serious inequality issues. But I think there’s a certainly an argument for means testing all benefits.

Peter Whiteford
Peter Whiteford
13 years ago

BG, I’m not sure that I would dignify this with the name of science!

My other comments are a bit of a ramble across the issues raised by you and others. First, like a number of others, it is not clear to me that the divergence in patterns of family work that you describe – among two-earner families – are necessarily one of the major causes of trends in inequality. I think it very likely that they have contributed to some of the widening, but I dont know how much.

Bettina Cass and Mary-Anne OLoughlin did a paper a long time ago about the rise in the number of two-earner families and its implications for income distribution. They pointed out that if you move from a world of single earner couples to one of two-earners, if there is assortative mating you should end up with the same level of inequality as you started with, assuming that wives all earn the same percentage of their husbands incomes be that 40, 50, 90 or 120% of the husbands income. But as you point out this is not what has actually happened. What we have is an incomplete process.

However, to me it seems that the rise in lone parenthood and also in the share of completely jobless households with and without children) is likely to have been more important as drivers of inequality.

There is a question of interpretation that you referred to. In the past, a lot of inequality may have been hidden within households, but now it is more out in the open. In addition, Australia has a relatively high aggregate employment rate but also one of the highest shares of completely jobless households. My interpretation of this is that we have a system that financially allows people to live in jobless households. If you compare Australia with Southern Europe in particular there are many more multi-generation households, so the jobless live with their parents or grand-parents. This narrows the distribution of household income but also makes the transfer system look less progressive!

Finally, it is worth noting that it appears that since the 1980s Australia has experienced a large increase in inequality of market incomes of the order of 4-5 percentage points, or about 10% relatively, but little change in disposable income inequality your welfare state at work.

Having said this even though all policy wonks concur that it is disposable income inequality that counts, maybe a lot of the population think it is market income inequality that is most important. What we are doing is saying that an increase in the real value of the age pension or in family payments offsets changes in the share of jobless households or in the number of lone parents. Strictly speaking they do, but someone who is single long-term unemployed may not feel the improvements.

Peter Whiteford
Peter Whiteford
13 years ago

NPOV

At the moment, the bottom 50% of the Australian households receive 80% of the cash transfers (this doesn’t include health and education), so it can be conveniently calculated that if you wanted to redistribute within the existing welfare budget alone you would probably need to get rid of all payments to people above the median household income which in 2003-04 was les than $30,000 (equivalised)

Peter Whiteford
Peter Whiteford
13 years ago

Please note that the implication of my comment 24 is that this is the redistribution within the existing social security budget which is comparable to the increase in taxes and transfers calculated as necessary to get the Gini coefficient down from 0.30 to 0.28. Whether it would actually do this, I can’t say from the data available

Anthony
Anthony
13 years ago

“even though all policy wonks concur that it is disposable income inequality that counts, maybe a lot of the population think it is market income inequality that is most important”

I suspect there’s an element of truth in this. ‘Policy wonks’ presume that as long as people are having their household incomes somewhat equalised, it doesnt matter where the money comes from. Yet is it really true that people value all money flowing into the household in the same way and that they make no distinction between earned income, a fathers tax credit and a mothers family payment? The fairness of earned income in particular is likely to be evaluated differently because it is a reward for individual effort. Adequacy is one thing, but arguably there exist important social norms of wage justice that link reward with effort, skill, training, and responsibility in a job. Unsurprisingly fairness of pay is often assessed by looking at the rewards given to people doing similar types of work. Accordingly, changes in wage relativities based on either the vagaries of the market or a workers bargaining power will typically be considered ‘unfair’, regardless of what the tax-transfer system does to equalise outcomes in terms of household disposable income

Backroom Girl
Backroom Girl
13 years ago

I’m sure there’s some truth in that, Anthony though the extent of unfairness you feel may also have some relationship to how you feel you are travelling yourself. I suspect that the reason many low-income workers feel somewhat resentful towards income support recipients is that they struggle themselves and they see another group whose standard of living is not dissimilar to their own who don’t have to work hard, possibly at jobs they don’t like. OTOH, I suspect most people who are doing OK themselves don’t really spend too much time feeling aggrieved about how much their boss earns or even their CEO.

But maybe people’s conception of fair reward has more to do with effort, rather than the type of job you or someone else does. I’m sure that many people in jobs with regulated wages have the experience of looking at the person in the desk next to them who doesn’t put in anywhere near the effort that they do but still gets paid the same. It’s pretty damn hard to be fair to all of the people, all of the time – all you can ever hope to achieve is something that most people think is reasonably fair most of the time, I reckon.

But I think a really useful observation of Peter Whiteford’s was that income inequality in Australia is probably just as much, if not more, a function of household joblessness as it is of CEO salaries. While the mnumber of jobless couple families has been falling here for quite some time, the number of jobless single parent families has only really begun to fall more recently. But he’s also right to observe that household joblessness here is in part evidence of the adequacy of our income support payments, rather than the reverse.

spog
spog
13 years ago

Anthony@14

I’m not sure I understand what you mean by the Gov’t will compensate people with kids for splitting up, but not those without.

People without kids who split up can get income support in the same way as people with kids. If you equivalise it, at maximum rates it’s probably roughly comparable in amount as well. Or did you mean something else?

Anthony
Anthony
13 years ago

Spog, I gather you’re referring to the effect of means testing, whereas I was referring to the question of eligibility for categories of income support. So it’s true, if you’re an unemployed person looking for work who happens to be partnered with an average income earner, you won’t be entitled to unemployment benefit, but if you separate from that partner you will be entitled to a fortnightly payment. This is the effect of the means test. The point is that before the split you were eligible for the payment, but the means test precluded any actual fortnightly payment, whereas after the split you’re still eligible for exactly the same payment, but now you’ll get it because of your reduced household income. So people without kids who split up can get income support i nthe same way they could have got income support while partnered except for the means test.

In contrast, if you’re part of a couple with kids and you split up and become the primary carer for the kids, then the government’s created a special income support category just for you.

My comment was a response to BG’s comment that the government shouldn’t compensate people for the adverse financial efffects of their decisions to split up. I pointed out the givernment, through sole parents payment, was trying to compensate kids for the adverse financial consequences of their parents’ decision to split up.

But, you’re right, how people decide to construct households will effect their entitlement to any given category of income support. It’s another moral hazard problem: does the means test (never mind sole parent payments) encourage couples to split up and form individual households?

spog
spog
13 years ago

I see what you mean now, Anthony, although I still don’t think you’re quite right. Before the split (and assuming the kids are young), a principal carer of a child in the couple is eligible for parenting payment, based on income and assets of the couple. After the split, the principal carer is eligible for parenting payment, based on their own income and assets. That’s exactly the same as “unemployment benefit” in that sense, and it’s been that way since the late 1990s.

Parenting payment is paid to a principal carer based on the fact they have young children. Like “unemployment benefit”, if you’re a member of a couple both incomes are counted; if you aren’t, only your own is used.

If the kids are older (eg, 8+ years), it’s “unemployment benefit” (Newstart Allowance) regardless.

Anthony
Anthony
13 years ago

Spog, you’re right that the introduction of aprenting payment in the mid-1990s blurred the existing categorical structure of welfare payments somewhat. But it doesn’t undermine my point that the government is more willing to subsidise households doing it tough when there are children involved than when the households are childless. And yes, I definitely wasn’t factoring in the more recent welfare-to-work reforms of the Howard government: these, I think, do represent a fundamental shift in the categorical payment system as it had evolced of the past century: that is, the government is now less willing to compensate a person with kids who is suffering the adverse financial consequences of a break up. I’m not sure whether BG would see this as a good development or not.

By the way, my mis-spelling ‘government’ in my last post as ‘givernment’ in a thread on welfare apyments seems a slip of an exquisitely freudian type.

spog
spog
13 years ago

Anthony, I didn’t notice it before, but I have to agree, “givernment” is brilliant.

NPOV
NPOV
13 years ago

BG, I don’t think I ever thought (or claimed) you were being “mean” to single parents, but I do question claims that there the financial discentives to repartner have much effect on actual behaviour. No-one denies that financial incentives are strong, but I really do start to wonder when it’s implied that they override far more basic needs such as that for love or companionship. When it’s claimed that the structure of welfare payments cause people to end relationships, or hesitate to form new ones, or give birth to children as single parents, or even fail to use adequate contraception (supposedly because in the heat of the moment lustful individuals know they don’t have to worry too much about the financical implications of accidental pregnancy), or any number of similar claims, I simply want to see the proof. It should be relatively straightforward to provide too, as we have a range of countries all over the world with different systems, some that are very generous to single parents, others at aren’t at all. If someone can show a strong correlation between actual behaviour and welfare structure, I’m all ears.

(FWIW, as an example of the sort of mean-spiritedness that the current welfare system can exhibit – many years ago when my mother was singlehandedly responsible for raising me and my two younger sisters, when it was determined that one of my sisters was overseas for a 3 week period, the family assistance office demanded that 3 week’s worth of benefits should be repaid – as though somehow having my sister out of the house for 3 weeks suddenly meant there were no expenses involved in maintaining a family of 3 children).

NPOV
NPOV
13 years ago

Sure, I can imagine that low-income earners might hesitate to move into together on the basis of reduced financial assistance. And if this is having a discernible impact on their ability to establish long-term stable relationships then there’s reason for concern.
I’m more bothered about people that make claims like “single parent payments and the baby bonus are encouraging women to get pregnant just so they can get the money (and then spend it plasma TVs)”. Now, I’m not a big supporter of the baby bonus particularly, but comments like that really don’t seem to be made out of any genuine concern for the facts.

Patrick
Patrick
13 years ago

Didn’t welfare reform in the late 90s in the US significantly reduce unemployed single parents?

spog
spog
13 years ago

I think BG would agree that the partnering disinentive she refers to is not all, or perhaps even mainly, about relationships not forming. There is a significant penalty attached to reporting said relationship to Centrelink, not forming it.

NPOV
NPOV
13 years ago

BG, then it would seem like generous single parent benefits should be an effective way to keep abortion rates down – except that Australia’s abortion rates are actually quite high compared to, say, most continental European nations.

Re reporting vs forming: it’s hard not to wonder whether Centrelink’s recent advertising blitz on the evils of not being entirely truthful when claiming benefits has led to otherwise happily cohabiting couples splitting up on the basis that the risk of being caught and the downside to reduced payments outweighed the pros of cohabitation.

trackback

[…] Don Arthur over at Club Troppo, and followed up by Andrew Norton, Andrew Leigh, Will Wilkinson and Backroom Girl. The idea that libertarians and ‘progressives’ could hammer out some of their […]