Unemployment & DSP

Good economic news today, with reported unemployment dropping from 5.8% to 5.7%. It is looking increasingly like I was right to be optimistic about the economy.

It’s just a shame that Rudd had to waste billions of dollars on impotent fiscal policy, it’s a double-shame that he will get the credit for a recovery that was always on the cards, and it’s a triple-shame that the hangover from the fiscal policy will lead to a slightly slower economic growth. Still, good news is good news, and low unemployment is worth celebrating.

But a work-mate just pointed out an interesting factoid.

Over the past quarter there has been a rapid increase in the number of people going on the Disability Support Pension (DSP). This makes sense. In a downturn, it becomes relatively more attractive to get DSP payments and so we are hit with an epidemic of “bad backs” and “depression”.

I decided to check the consequences of the “DSP mini-boom” on the unemployment number. DSP recipients don’t count towards unemployment… But without DSP, those people would have been on Newstart Allowance, and would be counted as unemployed.

The conclusion was that without the DSP mini-boom, the unemployment rate would still be 5.8%.

14 thoughts on “Unemployment & DSP

  1. The participation rate went from 65.1% to 65.2%. I’m not sure but I guess that is attributable to people going on DSP as well. The Aussie dollar absolutely loved it anyway.

  2. What is the margin of error in unemployment calculations? Could a 0.1% change just be noise?

    Ive also heard that DSP numbers rose precipitously during the mid 90s to mid 2000s as the government sought to ‘hide’ unemployment in pensions – something in the order of a fourfold increase in the number of DSP recipients. Never seen any actual evidence for that, I’d be interested to see if the facts actually bear it out.

    According to the 2010 budget (http://www.fahcsia.gov.au/about/publicationsarticles/corp/BudgetPAES/budget09_10/pension/Documents/Pension_Review_Report/appa.htm), we can expect to have 742,000 people on the DSP.

    We have a population of 22 million. Two million claim the aged pension. 1.25 million are unemployed. About half a million claim other pensions. 4.2 million are children under 15. Even exluding those not employed, and not claiming government benefits (stay at home mothers, the independently wealthy, self-funded retirees, etc.), I refuse to believe that one in 19 working-age Australians is medically incapable of work.

  3. Actually it’s worse than that. My good friend Ashton de Silva pointed me to the size of the standard errors. If the participation rate is out by 1se and the unemployment rate out by 1se, the overall unemployment rate could be 6.4%.

  4. That seems too high.

    According to the ABS LFS Bulletin (p.33), the SE for the unemployment rate is 0.1 and for the participation rate is 0.2

    For September, the trend PR was reported as 65.1, so if we assume actual PR was 65.3, that pushes the trend UR from 5.8% to 6.0%. Then adding the 1se for UR yields a figure of 6.1%, as compared to your 6.4%.

  5. “But without DSP, those people would have been on Newstart Allowance, and would be counted as unemployed.”
    .
    The alternative way to view this is that it isn’t such a bad thing if the cost of Newstart, with all the various waste-of-time and hassle the recipient programs, happens to be more than the DSP. I really find it hard to imagine there are too many people on Newstart or some DSP (who should really be on Newstart) that don’t want a job, so if there is no difference in them getting one compared to if they get a DSP, then the cheaper one is better.
    .
    “I refuse to believe that one in 19 working-age Australians is medically incapable of work.”
    .
    You can do that easily because 742,000 divided by 22 million gives you 3.3%, not 5.3%.
    .
    Personally, I don’t find it very hard to imagine that 3.3% are medically unfit to work, especially with an aging population. You can already subtract 1% for schizophrenics who can never hold down jobs, which gets you to 2.2%. You can probably subtract another 1% (at least) for various other mental health conditions — no doubt some caused at least in part by lack of work (e.g., depression) and others caused by things like strokes so people really are incapacitated.

  6. “You can do that easily because 742,000 divided by 22 million gives you 3.3%, not 5.3%.”

    Actually, I just realised that we both have it wrong. “Working age” is population minus those over 65 (using aged pension as a proxy), minus those under 15. So its 22-2-4.2=15.8 million.

    15,800,000/742,000 = one in 21, or ~4.7%.

    The aging population argument doesnt go to far, because the old are largely on the aged pension, not the DSP.

  7. “The aging population argument doesnt go to far, because the old are largely on the aged pension, not the DSP.”
    .
    I guess what we really need is the data (the link doesn’t seem to be correct — I will look for the report if I get some time — it would be nice to know what people are actually claiming). I was largely thinking of people 40+ who are overweight/obese and have been for some time, and have various associated problems with it. Health outcomes for people who are obese awful, and given the rate has been rising steadily for quite some time to quite surprising levels, that certainly must contribute.
    .
    It would also be interesting to know what the long term effects of some of the more modern drugs is (notably ecstasy) which most probably do have long term consequences such as increased rates of depression. That would probably add a few more people into the relatively useless category.

  8. “the link doesn’t seem to be correct — I will look for the report if I get some time — it would be nice to know what people are actually claiming”

    For some reason it has picked up a “)”. If you copy and paste the link, it should work fine – sorry about that.

    “It would also be interesting to know what the long term effects of some of the more modern drugs is (notably ecstasy) which most probably do have long term consequences such as increased rates of depression.”

    It would be interesting to see, but I expect that it would be dwarfed by old-fashioned alcohol-related health problems.

  9. The 6.4% calculation
    According to the ABS release (Cat. no. 6206 pp. 33)

    • The standard error of Persons Employed is 40.3 (000)
    • The standard error of Labour Force is 41.1 (000)

    According to page 7 of the same release:
    • 10805.6 (000) Persons are employed are employed
    • 11464.3 (000) people make up the Labour Force.

    Adding the standard error (41.1) to the Labour Force results in an adjusted Labour Force estimate of 11505.4 (000).

    Subtracting the standard error (40.3) from the persons employed number (11805.6) results in an adjusted Employed Persons count of 10765.3.

    By subtracting the adjusted Persons employed value from the Adjusted Labour force value, then dividing it by adjusted Labour Force value and multiplying it by 100 yields an unemployment rate of 6.4%.

  10. Thanks SD. That makes sense.
    .
    “It would be interesting to see, but I expect that it would be dwarfed by old-fashioned alcohol-related health problems.”
    .
    I’ve no doubt that’s true, but I was thinking of things that would cause the _increase_, rather than sustain the overall rates (like alcohol).
    .
    For example, lets say 20% of 30 year olds took a lot of ecstasy when they were young, and that 5% of them now suffer problems from it compared to a baseline of zero (too bad they won’t smoking dope, which has essentially no long term cognitive effects once you give up). This means that you get a 1% increase in the number of unemployable people (i.e., 20 * .05), so what are minimal effects on the average individual end up pretty large.
    .
    The figures would look even worse for the obese, who now make up almost 20% of the population. If a mere 20% of these guys have chronic health problems (like bad backs etc.), then you get 4% of people who are not going to be especially fit for work (or anything else). You could add just being overweight too to those numbers.
    .
    These are just speculations on my part — but it’s just not clear to me what the real rate of malingering is.

  11. Sinclair,

    Thanks, that makes sense.

    The standard errors I used probably yielded a different figure to your calculation largely because of rounding. A published SE of 0.1pp could actually be anywhere between 0.14pp and 0.05pp.

  12. “I refuse to believe that one in 19 working-age Australians is medically incapable of work.”

    Being incapable of work isn’t a qualification requirement for DSP. Broadly speaking, what’s needed is a measured functional loss of at least 20% which results in the person not being able to work for at least 15 hours a week at normal wages. The functional loss is assessed using tables in a schedule to the Social Security Act 1991. Before 1 July 2006 the requirement was not being able to work at least 30 hours. That’s a lot of work really.

    If you stop thinking about DSP as being for “invalids” who can’t work, and instead being for people who can only work part-time, the numbers are easier to understand. Of course, not everyone thinks this is an appropriate setting for DSP eligibility…

Comments are closed.