Slicing the Pie


I’ve been playing around recently with the new income distribution data that the ABS has released. I’ve created the graphs above, which show the change in real hourly wages and real income across the distribution, over the first eight years of the Howard Government (1995-96 to 2003-04).

  • Hourly wages rose by 6% at the 10th percentile, 13% at the median, and 16% at the 90th percentile.
  • Disposable family income rose by 7% at the 10th percentile, 16% at the median, and 17% at the 90th percentile. 

In both cases, the bottom few percentiles appear to have fallen quite substantially (down 20%). But given questions about the accuracy of the ABS’s sampling for the very poor, I haven’t shown those big drops.

I’m not quite sure what to do with this, but perhaps some readers have ideas.

Update: John Quiggin speculates about possible causes.

This entry was posted in Inequality. Bookmark the permalink.

14 Responses to Slicing the Pie

  1. Corin says:

    Andrew – I take it the graphs are an annual growth rate. Also does this account for people getting older and moving up the scale or even into retirement? i.e. It’s not a ‘personal’ map and I take it to mean if I leave Uni and move from part time work at say the 10th percentile and become a lawyer the next year at the 40th percentile this won’t be taken into account on an individual basis?? I assume it is an average of all people – so social mobility won’t be mapped as well? Simply across the average of all people what’s the real growth in wages at each percentile??

    Interesting stuff though. I think if you can consider it in alignment with increasing or decreasing social mobility it would be very useful in arguments about relative levels of equality across the decade. As you suggest in Imagining Oz – Inequality still matters even if we are all getting a bit richer??

  2. cam says:

    Why is the 40 percentile lower in income growth than the 25 and 60 percentiles? An artefact of the data collection/collation or something real?

  3. Sinclair Davidson says:

    At the very least, you could write it up as a Lies and Stats column (nice piece in the Oz today – the teachers union won’t be pleased). I don’t know if you can match the data but two ideas:
    1. the link between unionisation and inequality. If we assume low-income earners are more likely to be unionised, the their wage growth has declined as union membership has declined. Alternativily, wage growth has been lowest where unions are more active.
    2. A criticism of tax ‘reform’ has been that high income earners have benefited more than low income earners. (I imagine the data you have is pre-tax). The data seems to show that wage growth has been highest where the impact of tax ‘reform’ has been highest – i.e. by cutting taxes on the ‘rich’ the government has boosted productiity among high-income earners. (As you know this is a good thing as it promotes the ‘trickle down effect’).

    I don’t know if you can data match, but ideas are not mutually exclusive and could form one big paper.

  4. Andrew Leigh says:

    Christine, the change is from 1995-96 to 2003-04 (implying real family disposable income growth of 2% per year at the median). Your point about changing demographics should matter in the long-run (Jeff Williamson shows that “fat cohorts” get low returns), but perhaps not so much in the short run.

    Cam, I don’t know what’s going on with hourly wages. Maybe you can tell a minimum wage story at the 20th percentile? Just speculating, though.

    Sinc, the hourly wages are pre-tax, disposable family income is post-tax. I like the idea of looking at the effect of tax changes (indeed, I have a piece on this in the first issue of Curtin Uni’s new “Public Policy” journal). I’ll mull on it more. As to union status, it’s not asked in the SIHC – so rather heroic assumptions are required.

  5. Sinclair Davidson says:

    Is there a working paper version of ‘political economy of tax reform’?

  6. derrida derider says:

    It’s important to remember that the great bulk of those households in the bottom vingtile are dependent on social security, which is very badly measured in the income surveys (though it’s true some of the mismeasurement is a fixed error that washes out when comparing income sureveys over time). It’s not just the lowest percentiles where the data is really dodgy – that whole lowest vingtile should be treated with suspicion.

    And earnings data which implies very low hourly rates of pay should also be treated with suspicion. The stated income of the self-employed in these surveys is especially meaningless, as respondents usually don’t know it and it’s anyway very sensitive to exact definition.

    It’s interesting, f’rinstance, that in the HES the bottom decile of household income has higher weekly expenditure than the third decile.

  7. Matt Cowgill says:

    “It’s interesting, f’rinstance, that in the HES the bottom decile of household income has higher weekly expenditure than the third decile”
    that reminds me of this:

  8. Andrew Leigh says:

    DD, it is true that the data aren’t perfect. But when it comes to income distribution, these surveys are the best we’ve got, and you have to stretch a bit to explain why the changes in the above charts are all driven by measurement error.

    My interpretation would be that, using the best available data, the first 8 years of the Howard Government saw rich families do better than middle-class families, and middle-class families do better than poor families.

  9. Pingback: John Quiggin » Blog Archive » Inequality on the rise

  10. Fred Argy says:

    Andrew, your figures are most interesting but leave me confused. We all knew that wage dispersion (market inequality) has been tending to increase and this is confirmed by your figures. (I suspect the trend will accelerate as a result of the new workchoice legislation).

    That part is understandable. But your household disposable income figures are puzzling. The ABS GINI coefficient has shown no significant variation over the period 1995-6 to 2003-4. Yet your figures on disposable family income (which I presume include social security transfer payments such as Family Benefits) suggest that income inequality has been increasing in that period. I know the GINI has limitations but are the two sets of figures compatible?

    The story I have been telling is that (under Howard) “inequality of market incomes continued to rise on a trend basis but this effect was largely neutralised by greatly improved family and other social benefits” (I am quoting my Australian Institute discussion paper p. 17). Was I wrong?

  11. taust says:

    As one who experiences the benefits of so called shortages of skilled workers. (Albeit in an industry that is for all practical puposes fully open to foreign competition in all factors). Is there any long tem trend data on the long and short term inward and outward migration of skilled workers.
    I would guess right now we have a net inflow of both expatriate australians returning and of “foreign” workers coming in. Reversing the trend for the ten previous years.
    It would be interesting if the data series covered the 1980″s the last time there was a skill shortage scare because of the resources boom (on the back of the boom Frazer won an election and then, like the rest of us, p****d the gains up against the wall).

  12. Andrew Leigh says:

    Fred, I’m also confused as to why my figures don’t line up with the ABS numbers. I didn’t make any adjustment for household size, but that couldn’t be a full explanation. I’ll let you know if I come up with a better explanation.

    Taust, the data on migration is frustratingly imprecise (it’s based on those little cards we fill out each time we pass through Customs). Graeme Hugo has done a lot of work on the issue, and a 2004 Lowy Institute report by Michael Fullilove and Chloe Flutter had some good data too.

  13. Ben says:

    Apologies if this is vaguely off topic, but there was an interesting snippet on Radio National’s Saturday extra last week.

    Andrew West spoke about his book “Inside the Lifestyles of the Rich and Tasteful”. He divides the Upper Middle Class into the “Materialists” (libertarian outlook, catchcry = FREEDOM) and the “Culturists” (communitarian outlook, catchcry =AUTHENTICITY). Interesting stuff. Downloadable MP3, interview around the 1 hr 5 min mark.

    [audio src="" /]

  14. Pingback: Andrew Leigh » Blog Archive » What they’re saying about us in Luxembourg

Comments are closed.