A topic that the New York Times visits from time to time is that of income inequality. In the United States, the gap between the highest and lowest earners has been increasing over the last 80 years or so. A recent article returns to this theme and provides further insight into the trend. It cites research from the new book “Unequal Democracy” by Larry M. Bartels, which indicates that income inequality has increased far more under Republican presidents than under Democrats.
The table below shows average annual changes in income in the US from 1948 to 2005 across a spectrum of wage earners. (This is done by dividing the data into percentiles; for example, 40% of the population earns less than the 40th percentile wage and 60% earn more). Over this period, under Republican presidents high income earners (95th percentile) saw their income grow an average of almost 2% each year, while the wages of low income earners (20th percentile) grew by less than 0.5% each year. In contrast, under Democrat presidents, there was much less variation across the income spectrum and low income earners in fact saw slightly higher rates of income growth than high income earners. Under Republicans the income gap widened, while under Democrats it tended to contract.
| Percentile | Under Democratic Presidents |
Under Republican Presidents |
| 20th | 2.64% | 0.43% |
| 40th | 2.46% | 0.80% |
| 60th | 2.47% | 1.13% |
| 80th | 2.38% | 1.39% |
| 95th | 2.12% | 1.90% |
Annual Average US Income Growth 1948-2005*
This inspired me to attempt a similar analysis (with a bit of prodding from a regular Mule reader). Unfortunately, the Australian Bureau of Statistics (ABS) have only surveyed income distribution over an 11 year period from 1995-2006. Over all but the first year of this period, the Liberal/National Party coalition was in power and John Howard was Prime Minister. This makes a comparison of income inequality under Coalition and Labor Governments impossible. However, it does allow us to see whether or not income inequality increased under the recent Coalition Government as it did in the US under Republican presidents. The table below shows the average growth** in real weekly earnings of Australians (where “real” means adjusted for inflation). Income data is taken from the ABS Household Income and Income Distribution survey, and the inflation adjustment used the ABS Consumer Price Index (CPI) to express wages in terms of their equivalent in 2006 dollars.
| Percentile | Average Growth |
| 10th | -0.48% |
| 20th | -0.13% |
| 30th | 0.08% |
| 40th | 0.16% |
| 50th | -0.09% |
| 60th | -0.13% |
| 70th | -0.16% |
| 80th | -0.14% |
| 90th | -0.15% |
Australian Weekly Income 1995-2006*
While this period was a time of strong economic growth for Australia, wages failed to keep up with inflation across most of the income spectrum (only the 30th and 40th percentiles saw their wages increase in real terms). While the lowest earners (10th percentile) did see the biggest decline in real earnings, there was in fact very little variation across the percentiles. Below is a graphical representation of real wages for each of the percentiles, which reinforces a picture of very stable wages across the population over the course of the Howard years.
Australian Income Percentiles 1995-2006*
While the data is too limited to make conclusions with much statistical significance, it does appear that the Coalition, our (rough) equivalent to the Republicans, have not pushed up income inequality in Australia.
If anyone knows of a longer data series, let me know and I’ll extend the analysis.
* All figures are adjusted for inflation.
** The figures are compound annual growth rates.
Possibly Related Posts (automatically generated):
- The Mule trips up (21 April 2010)
- Australian Property Prices (30 June 2009)
- Junk Charts #3 – US Business Lending (23 February 2010)
- The Mule on Mortgages (13 February 2010)
{ 1 trackback }
{ 12 comments… read them below or add one }
There is an unnecessary focus on income inequality as a reliable marker of dysfunction in a society in the predominantly leftist media (and you do quote the New York Times), and I think this has been accepted a little uncritically
I think a better index to focus on is the absolute growth in wages as determined by purchasing power during the relevant admininistrations.
Or – How many hours of labour does it take a person in the lowest decile of income to earn enough money to buy a loaf of bread?
Without a doubt this time has decreased across all capitalist economies over the past 100 years. (reference needed)
Has the rate of decrease in the number of hours labour required to buy a loaf of bread in the poorest decile decreased at a faster rate in right wing or left wing administrations across a range of separate countries’ economies?
I have read data (god knows where, probably in The Economist, because that’s all I usually read on this sort of subject) that suggests that right wing administrations do better in this regard. (better reference needed)
It may even be a necessary condition for faster improvements in living standards for the poorer deciles that there is an increasing disparity in earnings from the top and bottom earning deciles
Appeals for a more egalitarian society should be identified as the nonsense that they really are. It never worked for the communists. It ran down Australia, the UK and the US when they flirted with socialism in the pre Thatcher and Reagan era.
To rephrase my question, does it really matter to someone on a low income if there is a 50% increase in luxury yachts in Sydney Harbour if he can now afford to by a Hyundai Getz?
@Sean: John Quiggan has done some work on income inequality over time in Australia and has some data (via references) going back to 1975. This seems to show that inequality did increase between 1975 and 1996, although I’m not sure the resolution is sufficiently fine to tie the changes to the party in power. See: Globalisation, neoliberalism and inequality in Australia
@Tristam Smyth: You make a valid point as far as the US statistics are concerned. Since they don’t seem to be adjusted for inflation, we don’t know whether 0.43% growth under republicans actually allowed the lowest quintile to purchase more than the 2.64% growth under democrats. If inflation was consistently higher by 2.3% under democrats, then your way of asking the question would lead to the answer you expect. However, this doesn’t seem to be the case. In fact, it appears that inflation has been slightly higher under republicans (at least in the period 1962-2001); see:
http://pla.blogspot.com/2002_10_27_pla_archive.html#83853463
As far as the Australian figures go, they are CPI adjusted, so that the lowest decile clearly experienced an increase in “the number of hours labour required to buy a loaf of bread”. This was under a right wing government, and during a record-breaking global economic expansion and low domestic inflation, clearly contradicting your assertion that the lowest decile has seen consistent improvement in real incomes.
One thing that has been shown to be consistently different under Liberal and Labour governments in Australia (and similarly in the UK) is the suicide rate; it is higher under right wing governments.
See the paper Suicide and political regime in New South Wales and Australia during the 20th century or the reportage at:
http://www.abc.net.au/rn/talks/8.30/helthrpt/stories/s690232.htm
http://news.bbc.co.uk/2/hi/health/2263690.stm
Of course, perhaps suicide is also not “a reliable marker of dysfunction in a society”
No doubt, Mule, that this chart may well change drastically from 2005-2008. The uber-rich and their fairly rich minions are facing challenges as their ill-conceived investments become increasingly precarious. Where else to go? Looks like all asset bubbles are busting or busted (real estate, food, oil, etc.). Perhaps the day of the $400 million CEO has finally passed? And I know it’s out there somewhere, but I would like to see the stats on a large corp where they show the share dividend with the CEO compensation included (aka normal) and then taken down to a more normal level (aka should be normal, ie no more than $2 million/year). Anybody have that data?
@Tristam: As Mark notes, adjusting the figures for inflation accounts for purchasing power. In fact all the figures in the post are adjusted for inflation, including the US figures (which were taken from the table in the NYT article).
You do, however, raise a good question when you ask does it really matter. While the instinctive reaction of many people is that increasing income inequality is not fair, this is not necessarily a rational position. In an earlier post I referred to Dan Ariely’s excellent book Predictably Irrational. One of the points Dan makes is that humans are much better at making relative judgements than absolute (How much is a 4 cylinder car worth? Who knows, but it’s less than the 6-cylinder). As a result, although most of the 10th percentile of Australian income earners have food, clothes and shelter, unlike countless people around the world, they will tend to judge their circumstances in comparisons to other Australians. So, even though you can certainly argue that increasing income inequality shouldn’t matter, in practice is does tend to affect people’s happiness. Although, the solution may be for everyone to read a book like Affluenza rather than closing the gap.
@Bast: Although you’re right that the fortunes of many of the richest will have taken a downturn recently (some former Lehman Brothers employees, for example), it’s probably not something that you’d pick up in data like this: the 90th percentile is far too low. All the CEOs would be well and truly above the 99th percentile.
Your analysis seems at odds with St. Vinnies.
@David: thanks for the link. At first glance, that paper also uses ABS data, so I’ll look more closely to see why our conclusions differ.
I think a better indication of wealth distribution is ones ability to purchase a House, which has become an outright impossibility for the lower earners in society. As for our ability to buy a loaf of bread, yes it has become more manageable, but where does that loaf of bread come from? whose pocket is suffering to cause this saving?
Probably farmers?….. As its either been imported or Woolworths or coles have driven down their buying prices, which means the trade deficit increases and or working class jobs disappear. As for other essentials, such as electricity bills, water and so on, these have also risen in conjunction with or above wage inflation. Therefore I would have to postulate that the rich poor divide is indeed growing.
New post on the updated ABS coming soon. Here is a preview.
Hi Stubborn Mule,
Interesting subject. Allow me a couple of observations:
First observation:
I haven’t read Bartels’ “Unequal Democracy”, so I don’t know which sources were used in the book. Judging by the NYT article, where the table above appears, it seems the data comes from the US Census Bureau.
If so, then I suspect the data is part of the American Community Surveys. These studies, as their name indicates, are based on samples. If a sample fails to include extremely wealthy individuals (say, Bill Gates or Warren Buffett), the results (especially those of the upper echelons in the income scale) will be underestimated.
But, perhaps much more importantly, the American Community Surveys do not include capital gains as part of their income definition; while I believe it does include all government transfers (food stamps, unemployment, and such).
In this the ABS’s Survey of Income and Housing is quite similar: it is also based on a sample and does not include capital gains (although it does include profit/loss from unincorporated business and net investment income, plus government transfers, i.e. pensions and allowances).
Why they exclude capital gains, while including most other forms of regular income? I don’t really know for sure. But I suspect this distorts the final result, in both the US and Aussie cases, regardless of political party.
Second observation:
Does it really matter? I suggest a different way to look at this question: any attempt to reduce inequality will likely face enormous political opposition. At least, this has been the historical experience. This, I would think, suggests a part of the population cares about maintaining high inequality levels.
Why should those at the bottom of the scale be indifferent?
Besides, this situation did not simply arise, for no particular reason. I suggest that its causes may conceivably be quite relevant.
I am sorry for the self-promotion, but I am writing some material about this topic and I hope to be posting it in my own blog shortly. If you are interested, let me know and I will give you my URL.
Cheers,
Marco
Marco: Omitting capital gains does seem to introduce an unnecessary skew in the results. You are right as well that there will always be estimation errors in sample-based statistics. Fortunately, however, the statistics here are quantiles which are more robust (i.e. less sensitive to inclusion/exclusion of outliers) than statistics such as the mean.
By all means post the URL to your blog post.
The link is
Magpie’s Asymmetric Warfare
http://aussiemagpie.blogspot.com/
I don’t have much stuff yet, though.
Marco
Hi, Stubborn
Happy new year.
As you are interested in income and wealth distributions, I’ve posted a short comment on the subject. It’s called “La Vie en Rose: the Irvine Index”. You can find it at:
http://aussiemagpie.blogspot.com/
I am still studying some material, so a more elaborate article is cooking.