With the “Occupy X” rallies gathering momentum and attention, inequality is suddenly a prominent political issue. It’s pretty clear what people in the US are angry about – their unemployment rate is high, growth prospects are low, inequality is high and rising. Business Insider has a good summary of America’s economic and social problems, in chart form.
It’s important to bear in mind that Australia is not the US. Inequality here is lower than in the US, social mobility is much higher; the unemployment rate is much lower. Whereas real median wages in the US have stagnated for decades, we’ve seen fairly solid real income growth across the income distribution. I think that Australians who are concerned about rising inequality should be aware of the facts, so as to avoid making overblown and unsubstantiated claims. It’s easy to dismiss people’s arguments if they’re based on a misunderstanding of the facts.
Below, I set out some facts about income inequality (as distinct from wealth inequality, which is quite a bit higher than income inequality in Australia and elsewhere).
The share of income going to the top 1% in Australia has doubled in the past 30 years or so. [fn1]
According to the ATO’s tax statistics, if you earned $248 192 or more in 2008-09, you were part of the top 1%. The average income for an individual in the top 1% of taxpayers was $554 185. The median income for that year was $45 200, so the average income for someone in the top 1% was more than 12 times the median.
Top 1% income share in Australia
This is a pattern common to many developed countries – a ‘great compression’ in the three decades or so following World War 2, and then a ‘great divergence’ beginning in the late 1970s or early 1980s.
Even though the pattern is the same, the extent of the compression and divergence has been quite different across countries. The US is the extreme example. At the end of WW2, their top 1% had about the same proportion of total income as Australia’s top 1%, but our great compression was more pronounced and lasted longer than theirs. Our great divergence has also been less severe. Our top 1% has increased its share of income from around 5% to around 10% in the past 30 years, whereas the top 1% in America has increased its share from about 8% in 1981 to around 18% in 2007. Note that figures which include capital gains put the 1% share in the US closer to 25%.
Top 1% income share in Australia and the US
When you add in the other OECD countries, it looks like Australia’s top 1% income share is around the middle of the pack. It’s important to bear in mind here that these figures relate to pre-tax income; countries that have a similar pre-tax distribution but differing degrees of redistribution through the tax and transfer systems will end up with very different levels of inequality in net incomes.
Top 1% income share in OECD countries since WW2
It’s a bit hard to figure out which country is which in that tangle of coloured spaghetti, so the chart below shows the top 1% income share for selected OECD countries in 2005. I chose 2005 simply because that’s the most recent year for which the World Top Incomes Database has information for more than a handful of countries. Again, this shows Australia as middle of the pack among developed nations for the share of gross income going to the top 1%. [UPDATE: A helpful reader pointed out that Norway's top 1% share spiked in 2005 due to tax changes in 2006, as can be seen in the chart above].
Top 1% income share in selected OECD countries in 2005
Of course, the top 1% income share is a very interesting piece of information, but it doesn’t tell us anything about the distribution of income among the rest of the 99% of the population. To get a more complete picture of income inequality, we need to use a measure like the 90:10 ratio or the Gini coefficient. A Gini of 0 implies perfect equality of incomes (everyone has the same income), whereas a Gini of 1 implies perfect inequality (one household has all the income). The closer a country is to a Gini of 1, the less equal its distribution. The data here are for net incomes, after taxes and transfers, and are based on household income, adjusted for household size. [fn2]
Measured using the Gini, our level of inequality (0.3) in the mid-2000s was slightly below the OECD average (0.31), but in a similar bracket to most European countries. Our level of inequality was below the level of most other English-speaking countries, and significantly below the United States (0.38). Although our income tax system has become less redistributive in the past decade or so, the tax and transfer systems as a whole still reduce inequality quite a bit in Australia.
Gini coefficient in OECD countries in the mid-2000s
Although inequality in Australia was middle-of-the-pack in the mid-2000s, the Gini in Australia has been drifting upwards. It rose from from around 0.29 in the mid-90s to around 0.33 in 2009-10. Between 2003-04 and 2009-10, the Gini in Australia increased by 0.022 points, a 7.2% increase in inequality.
Over this period, the share of household income going to the top 20% of households increased from 38.4% to 40.2%, after reaching 41% in 2007-08. In 2003-04, a household at the 90th percentile of the income distribution had 3.87 times the income of a household at the tenth percentile. In 2009-10 this ratio was 4.21 times. This increase in inequality has happened during the mining boom period, and also during a period in which the income tax system was flattened out a fair bit. It’s likely that both factors have contributed to the rise in income inequality in the mid-to-late 2000s.
The OECD inequality data currently don’t go beyond the mid-2000s, so it’s difficult to say whether other countries have also become less equal over the same period.
Gini coefficient in Australia – mid 1990s to late 2000s
One of the most biggest problems associated with higher inequality is lower social mobility – the longer the ladder between top and bottom, the harder it is to make it to the top. Researchers measure this mobility as the relationship between an individual’s earnings and his or her parents’ earnings. If this is 1.0, then your income can be predicted with 100% accuracy based on what your parents earn. This implies no social mobility. If the figure is 0, then your earnings have no relationship to your parents’ earnings, implying high mobility. Australia has a moderate level of social mobility, according to research by Andrew Leigh, with an intergenerational elasticity of earnings in the range of 0.2 to 0.3. Dr Leigh explains:
In other words, a 10% increase in a father’s earnings translates into a 2-3% increase in his son’s earnings. In terms of international rankings, this suggests that Australia is more socially mobile than the US (IGE=0.4-0.6), less socially mobile than Scandinavia (IGE=about 0.2), and approximately as socially mobile as the UK. Given that’s where we fit in the inequality rankings, this feels about right to me.
I’ve written before about my belief that inequality matters. Even if we manage to eliminate absolute poverty within a country, inequality can still have pernicious effects on society and on individuals. One of the most harmful potential effects is that, over time, large inequalities of outcomes can undermine equality of opportunity, making it less likely that hard working, smart kids from poorer backgrounds will have a chance to achieve their potential. However, it’s important to keep the debate in perspective, at least as far as Australia is concerned. Income inequality here is rising, but it’s still relatively moderate. We should be vigilant to make sure that that our relatively egalitarian society stays that way.
NOTE: An earlier version of this post used a chart from the OECD to suggest that Australia’s level of intergenerational social mobility is relatively high. The OECD report used underlying data from a working paper by Andrew Leigh. The final version of his paper had a different finding about the level of mobility, but unfortunately the OECD report remains online with the original chart intact. I have updated my post to take account of this. Dr Leigh explains the whole situation here.
[fn1] The charts on the income shares of the top 1% are based on data from the World Top Incomes Database.
[fn2] Equivalised household disposable income is used for the OECD Gini data.
Hi Matt, Thanks for your ever timely and expert presentation of the relevant figures. I’ve been wondering about these figures for Aus for the last week or so with the increasing coverage ot OWS.
Cheers, thanks for the comment!
Matt,
This is very interesting to look at the figures, but you’re only looking at inequality from an income perspective. Granted every person who has attended the various Occupy events worldwide has their own individual reasons for doing so, but I think income inequality can only be seen as one of them. Looking at it from this perspective could be a little too simplistic, I feel.
Hi Weh Yeoh,
I agree that income inequality doesn’t fully capture economic inequality. Wealth inequality is important too, as I noted in the post. Dynamics (change over time for particular individuals) are also important, but I have left out that dimension in this post. The study and measurement of inequality is a big topic and I can’t possibly do justice to it in one, relatively brief, blog post.
Nice one dude. I have a preference for 80:20, 90:50 ratios etc over the Gini coefficient, but you’re right on the money.
Worth noting the ABS have revised their Gini figures several times (changes in sample size and increased coverage of the defition of ‘income’) – with the revisions illustrated in Chart 1 of Peter Whiteford’s recent Troppo post: http://clubtroppo.com.au/2011/09/30/inequality-in-australia-%E2%80%93-are-the-rich-getting-richer-and-the-poor-poorer/. Of course, the upward trend is still obvious – just the exact timing, causes and size of the increase is harder to pin down.
I’m also skeptical of international comparisons of Gini coefficients – it’s a very sensitive instrument and I suspect the differences in the 0.28-0.32 range would disappear with a few sampling tweaks or minor changes to definitions. Again though, Australia’s position (firmly in the middle of the pack) is in no doubt.
Cheers Aaron.
I generally prefer ratios (90:10, etc.) over the Gini, too.
I’d love to see the 99:50 ratio in equivalised household disposable income over time, but you’d need the SIH CURF for that and probably couldn’t go back very far (1993?).
See Siminksi et el….explains methodology problems, changes in methodology over time by ABS – may go back earlier but not much.’Assessing the Quality and Inter-Temporal Comparability of ABS Household Income Distribution Survey Data’ 2003
Hey Matt, do you understand the Occupy X slogan of “We Are The 99%” to mean that they are protesting the wealth and income of the top 1% of the country that they are in? I saw a poster that linked the saying to the top 1% of the population owning almost half the world’s wealth. After Googling that information, it seems to come from a report from the UNU-WIDER, which stated that the top 1% of adults owned 40% of the world’s assets in 2000. If the groups are targeting global rather than national inequality, how does that affect your view of their political tactics?
Hi,
In short, yes, I took it literally and assumed that it meant they were protesting the wealth and income of the top 1% of their country.
Many, if not most, of the protestors would themselves be in the top 1% of the global income distribution, so the “we are the 99%” slogan wouldn’t make sense other than as a metaphorical, “we are all Spartacus” type of solidarity thing.
I nearly included this graph in my post: http://economix.blogs.nytimes.com/2011/01/31/the-haves-and-the-have-nots/. It shows that the poorest Americans have higher incomes than most Indians.
I’m not sure it does affect my view of their political tactics, I don’t think I follow your question
That’s a really interesting article in the New York Times.
I’m not sure how accurate it is, but I played around with the “how rich are you?” tool on http://www.globalrichlist.com/ until I got 1% at US$47500 (so roughly AU$46700). I think that’s about the median income in Australia (?), so it’s not inconceivable that most of the protesters would be below the 1% line in income, is it? Or perhaps I made a mistake.
I might have a stereotyped view of what protesters might be like – but also, it’s more than I earn. :)
Lucy,
Let’s take those numbers as given and assume that the 99th percentile cutoff for global income is approximately equal to the median income in Australia.
Let’s also assume that all the protesters have incomes below the Australian median.
That would still leave the protesters in the top 2-3% in global incomes.
What I’m wondering is, if the protestors are speaking on behalf of the poorest people in the world, rather than as the poorest people in the world, would that turn your opinion of their tactics from “let’s not exaggerate and discredit ourselves” to “you guys are dumb and you’re making it harder for first world citizens get perspective on world poverty”?
At highest point in 1950 due to wool spike – the top 1% commanded just under 15%, yet top 10% commanded 40%. Lets not get too obsessed with the top 1%. Top ten percent also of interest here as a source of inequality.
Lets not listen too much to politicians and the media (given politicians are approaching / or in the top ten percent and the media – well we know all about Rupert Murdochs power and business interests now. His empire has done well enough from pushing the politics of inequality and low taxes on the wealthy over the past three decades. His empire is a mjor source of the problem.
Its more than the top 1% versus the rest of us. Taxes really need to be raised at the top ten percent level (at least) to undo this sort of inequality, but yes – its clear something has to be done and if it means starting with redistributing away from the top 1% it needs to happen and fairly soon. There is too much social unrest, globally, and the majority needs to make a decent living without too much economic insecurity. If the rich wont trickle down, the poor will make them bleed down if it keeps going this way. Thats s sad fact of history. The 1% can never control the 99% unless the latter are happy enough to be controlled and they clearly are not happy, on a global scale.
Thanks Matt. I think the argument that things are worse in the US misses a few points. As the AACTU itself recognises the share of national income going to labour is at its lowest and that going to capital its highest (or almost) since records began to be kept. In addition we work long hours – about 45 per week. so we are working longer and longer to reward the rich parasites and their insatiable desire for more and more profit.
Further the idea that because we are not as bad as the US now we won’t be in the future – especially if China collapses – is I think a rather courageous view. we essentially have tow parties of neoliberlaism – the soft and hard versions battling it out.
This is also about having a say – the labor Government listens to the miners, the heads of the banks and the like, but not to you or me. I wrote something about this on my blog – some facts to support occupy Australia. http://enpassant.com.au/?p=11270
And the ACTU’s own figures show the richest 20% own over 60% of the wealth, while the poorest 20% own one percent of the wealth. And it is getting worse, as the recent ABS figures show.
What about the gender pay gap? As bad under Gillard as it was under Howard. There is much to be angry about.
Matt, nice post (and thanks for the double-plug). FYI, I have a longer timeseries on the top 1% available here. Hoping to get time soon to update it further still.
As to mobility, alas the OECD erred in using a draft version of the paper. Full story here. We’re in the middle of the pack for mobility.
Hi Andrew,
Thanks for that. I hope you don’t mind that I only cited your work indirectly, via the Top Incomes Database and the OECD!
I used your longer timeseries, updated using the 08-09 tax stats, to create the chart on p.24 here: http://www.actu.org.au/Images/Dynamic/attachments/7201/ACTU_Tax_Paper_4.pdf. Your series is very useful, thanks.
Thanks for alerting me to the problems with the mobility data, I’ll amend my post.
I thought that time series looked familiar Andrew.
Oh, and one other measure that we can produce for Australia is a <a href="http://andrewleigh.org/pdf/Australian_inequality.pdf"male income gini. Methodological issues aplenty, but it might go some way to answering commenters questions.
Andrew your link doesn’t work.
Matt, while decile ratios are easier to understand than Gini coefficients, the problem is that they do not capture any inequality caused by incomes above the top or below the bottom pont chosen. so for example, the 90/10 ratio effectively leaves out the incomes of the top 10% and the bottom 10%, which in some cases can make a big difference.
Hi Peter,
Yes, I suppose there is a trade off between simplicity and comprehensiveness. Measures like the Gini or the Atkinson and Theil indices tell us more about the entire distribution, but they’re harder to explain. I guess for the purposes of this blog I try to err on the side of simplicity, but that does have its costs.
Interesting post, Matt.
But many commentators on the OWS movement – and I know this wasn’t really the point of your post – miss one critical aspect of why the protests have spread across the globe. The generation that is currently taking to the streets is the most connected, globalised generation to grace this planet. It is as much about showing solidarity with counterparts in countries experiencing worse conditions than our own, as it is about protesting domestic socioeconomic circumstances.
With that said, I think the protesters here in Australia have done themselves somewhat of a disservice by adopting, almost verbatim, the language and rhetoric of the US protests. But then, this is hardly surprising when the root cause of inequality in Australia is the unquestioning adherence amongst our economic elite to distinctly American ideology.
Hi Matt, great post, but you mind including the Y axis description on the initial graphs. I can’t follow what this axis is representing.
Hi,
I take your point and will include Y-axis labels on future graphs, but I’m not going to re-do these. The chart titles double as the Y-axis descriptors.
The charts show the share of total individual income that was earned by the top 1% of income earners. The first chart shows the income share of the top 1% in Australia over time, the second adds the same measure for the US, the third adds other OECD countries.
Thanks Matt, I presume in millions of dollars?
Hi Daz,
No, they show the share (ie. percentage) of income going to the top 1%.
Matt,
I have a few observations about this analysis.
The WTID data for Australia are taxable incomes: thus, if I am not mistaken, they include government transfers (social security payments, dole, retirement payments). In other words, they overestimate market outcomes: they show a conservative view of inequality.
Further, as the WTID says (Warning section, Introduction page): “The rich, in particular, have a strong incentive to understate their taxable incomes”.
In view of these two things, I’d say that a correct interpretation of the data should consider the inequality they show as a best case scenario or a lower bound.
At least to me, this is not clear in your analysis, particularly when you suggest:
“It’s important to bear in mind here that these figures relate to pre-tax income; countries that have a similar pre-tax distribution but differing degrees of redistribution through the tax and transfer systems will end up with very different levels of inequality in net incomes.”
There is a further question: if one were to consider income taxes, why not consider other taxes payable by individuals? Think of GST.
But even if we do limit ourselves to income tax, in Australia the tax-free threshold was increased only once since 1996 (from $5.4K to $6K in 1999); while the top income threshold was increased several times, from $40K to $180K.
So, even though progressive taxation should reduce the inequality reflected in the TWID data, it is by no means clear to what extent it actually counters it (at least to me).
Another observation relates to international comparisons: the WTID data (which I believe you use in the third and fourth charts, entitled “Top 1% income share in OECD countries”) are not particularly suited to international comparisons: “Thirdly, the definition of income and the unit of observation (the individual vs. the family) vary across countries making comparability of levels across countries more difficult”.
Of course, the gist of your message is still pertinent: Australia is not the US. But then, again, it is not Denmark either.
Hi Matt,
Another question for you, on the theme of income tax and inequality…
Changes to the income tax schedule are obviously big decisions for a government. Peter Costello’s said a lot about his values, Wayne Swan’s say a lot about his. They are decisions that get reported widely. And they are about as hip pocket as it gets.
Yet the public appears to pay little attention.
Most voters, even politically interested ones, can’t describe any aspect of the recent changes to income tax, beyond the fact it has fallen. Even then, the majority of respondents to the 2007 and 2010 AES surveys said they thought taxes had “stayed the same” or “increased”, despite the cuts.
Do you think voters’ understanding of the system would be improved if employers were required to quote wages and salaries in ‘after tax’ figures? So that changes to the tax schedule were more obvious, and people knew what they were really getting?
There was a similar debate about prices and GST.
It’s hard to tell what effect that change would have.
But maybe a better understanding of what governments do might lead to a better alignment with what the public want. You make a good point in your tax report about the support for a more progressive system. Despite that, few people seemed to have noticed that Costello moved us away from that and that Swan has moved us closer again.
Nathan
I think you’re right that people don’t pay much attention. I think they would pay attention if there was a sudden shift in one direction or another. I don’t think people notice gradual, incremental change so much, even if (over time) those changes add up to something quite significant.
In my view, one thing that politicians can do if they want to make sure that their reforms are noticed is to make them simple and transparent. I think part of the reason that Swan hasn’t got as much credit as he should for some of the changes he’s made, particularly at the bottom end, is that LITO is an opaque, difficult to understand tax credit. Look at how much press the Government has (righly) received over the big boost to the statutory tax-free threshold; a lot of this analysis ignores the fact that it’s a smaller boost to the effective threshold (& thus a smaller tax cut than would be implied by the headline figure) because the higher threshold is replacing most of the LITO. So, simplicity counts. You get more bang for your political buck by making the effects of your reforms clear and easily understood.
I guess that’s the same point you’re making about employers quoting after-tax salaries. I think that probably would have a similar effect.
One thing that’s a mystery to me is that support for tax cuts has fallen, and support for increased social spending has risen, quite strongly since the early 1990s. Despite that, it’s still generally assumed that tax cuts are always and everywhere a vote-winning trump card that eclipses (say) spending on hospitals, schools, disability services, etc. I don’t know why that’s the case.
There’s a nice line about polling I once heard from Matthew Taylor:
“Polling is great. It tells you exactly what voters want. They want Swedish-style welfare and US-style taxes”.
I think that captures the mood in Australia at the moment. You listen to any election pitch, nobody is saying “We’ll deliver tax cuts instead of more services”. Tax cuts aren’t seen as an unstoppable trump card. Rather they’re saying “We’ll deliver tax cuts AND more services”.
[...] Inequality and the top 1% in Australia [...]
[...] Cowgill offers an excellent summary of Oz's top 1% on his We Are All Dead blogfor those interested in reviewing where we stand in the inequality [...]