Rob Bray points out to me some interesting data from 1999, in which Peter Saunders (SPRC, UNSW) asked respondents to place themselves in an income decile. Of course, 1/10th of the population falls in each decile, so if people are accurate, then the result should be 10 bars, each containing 1/10th of the population. But here’s what Saunders finds:
And for those who like decimal points with their data, here are the percentages:
|
Decile |
% |
|
1 |
1.3 |
|
2 |
4.8 |
|
3 |
14.1 |
|
4 |
23.7 |
|
5 |
25.8 |
|
6 |
15.6 |
|
7 |
10.6 |
|
8 |
3.3 |
|
9 |
0.7 |
|
10 |
0.1 |
In other words, only 1/10th of those in the poorest decile know it, and only 1/100th of the top decile are willing to admit it.
A decade on, it’d be worth seeing whether Australian are any better informed about their true position in the income distribution than we were in 1999.
Core Economics 
And why should we assume that the survey respondents are in fact evenly distributed across the deciles? Was there a dollar-amount income reported from each of them, to gauge the accuracy of their self-assessments?
Also, do most people know the income-level gradations corresponding to these deciles. Many might be surprised to learn that household income of more than $250,000 per year puts you in the top decile.
LikeLike
I guess it’s all relative. If you compare yourselves to your neighbours we would expect this kind of distribution (due to geographic convergence of similar households and incomes). How could you really expect people off the street who don’t have a keen interet in the subject to be able to compare themselves to people in other areas? I personally do have a keen interest, yet would have no idea about the typical incomes of regional towns and other captial cities. And yes, I would put myself in the middle decile.
I guess my interpretation is that this is partly a sign of guesswork (because we really don’t know) and partly a sign of comparison with people close to us.
LikeLike
As the other commentors have argued, it’s not clear that there is much information out there (about the cuttoffs for each decile) for respondees to work off. In light of that, reversion to the mean is a sensible way of ensuring you are not too wrong in your guess…
LikeLike
“Of course, 1/10th of the population falls in each decile” you say, rather casually. But I’d guess that most respondents did not understand what a “decile” means and so were answering a different question.
I know that I didn’t until you pointed it out. I may have come across the definition many years ago, but being outside the statistics field have rarely if ever seen it referred to since then, Indeed, you might say that the concept of decile is rather unnatural (for want of a better adjective).
Did respondents have the meaning of decile clearly explained to them? If not, they probably confused it with “income range” (in dollars) or something similar.
LikeLike
While there are clearly some issues around how people respond to this type of question, to my mind the most signficant feature is that some 70 per cent of people felt that they had below median incomes.
Such a perception may well suggest why so many Australian’s are vulnerable to lines such as a “Aussie Battlers” – and a belief that most people are doing better than them – and they don’t think this is fair.
Even those who think they are better off than the majority, generally only consider themselves to be a little bit better off.
This sort of data might also explain why people often see no contradiction between wanting lower taxes on themselves, but higher levels of government spending on services – in their minds there is a group of Australians much richer than them who are clearly well placed to pay more.
LikeLike
Andrew, what would be really interesting would be to have the incomes next to the deciles, so that your readers can work out where we fit (although I suspect I already know).
LikeLike
Interesting. At about the same time as this survey, I was working for a large and particular Australian organisation with an old-school clear and fixed pay scale. Every 4 years there would be an enormous staff survey conducted. One of the questions explained/defined and then asked responds what their total annual gross income was. By crosstabulating the answers to this question, with the answers to the preliminary questions on their rank/pay grade/time in, it was abundantly clear that a large number of staff spread across all ages and pay grades in the organisation (over 30% if I recall) self-reported an income that was lower than what we knew they were being paid.
Perhaps they were all negatively-geared tax ninjas, but there was other evidence to discount that.
Conclusion? there are lot of people who can’t seem to even quantify their own income; bit of a confound to asking them where that then fits into community norms.
In light of that experience would be interested to hear Andrew if you have the original paper, what did Saunders do to determine whether he really did actually have perfect distibution in his sample set – was it self-reported $$ income figures!
LikeLike
The complexity in this is that the decile bands are nowhere near equal size. Those around the middle are only $5-10K so it is very easy to be close(ish) in dollar terms but 2-3 deciles off.
The number of top 10% earners who don’t realise it is interesting . I think that is just due to environmental bias. e.g. More than 25% of people they know earn more than them therefore they can’t believe they are in the top 10%.
LikeLike
Does the fact that the survey responses resemble a bell curve mean anything? Or are Australian incomes not distributed in a bell curve? (My point being that the people might have been somewhat accurate about their position on the bell curve. This report on Australian incomes for 2007/08 at http://www.ausstats.abs.gov.au/Ausstats/subscriber.nsf/0/32F9145C3C78ABD3CA257617001939E1/$File/65230_2007-08.pdf shows a snail on page 11.)
I recently found out that I earn more than 80% of Australians. Which is great, but weird given that I am a junior public servant in my first couple of years out of uni. In other words, the majority of the public service, especially policymakers, are even more unrepresentative of the public than I thought we were.
LikeLike