Good data is vital in a crisis. What sort of quality are we getting from the Australian Bureau of Statistics (ABS)? The latest official unemployment figures were a big surprise, with a large fall. It slid to 6.8% — quite unexpectedly.
But a new dataset the ABS releases on weekly payrolls tells a different story. It shows the labour market weakening in August, a weakness that continued into September. The jobs index fell to 95.5, well below its pre-pandemic level of 100 on March 14.
Other data might also cause one to raise an eyebrow at the official unemployment number for August. The ranks of recipients of JobSeeker continued to swell in August, to a record 1.45 million.
Of course, statistical error is always possible. The official unemployment rate comes from a survey with a 95% confidence interval around the range 6.4% to 7.2%, meaning one time in 20 we can expect truth to be well beyond the reported number. Could that have happened this time? We await the next release to find out.
I am extremely concerned about the amount of underemployment in so many sectors across Australia and would like to see figures reflecting this. Technically, if one has a couple of hours work here and perhaps 3 hours work there a week – if one is lucky – you are in the workforce. However, one does not have a living wage – not even a modest one.
ABS unemployment data follows the international standard that specifies working one hour a week to be regarded as employed. That’s why there are also surveys of underemployment which are more useful.
The ILO definition is so vague that it doesn’t even meet the definition of being a definition.
As long as we continue to use meaningless figures like showing someone as employed by just doing one hours work it’s all a nonsense! If we halve it to 30 mins, we’ve halved the unemployment!! If we reduce it to 10 minutes, suddenly everyone’s employed!!!!!!
How good are statistics????
NATIONAL AFFAIRS ABS
makes employment figures bend over backwards
by Marcus L’Estrange marcuslestrange@yahoo.com.au 0450285163
News Weekly, May 30, 2020
April’s Australian Bureau of Statistics “Labour Force” unemployment figures of 6.2 per cent must be taken with a massive grain of salt. They are merely a political definition of unemployment, not an actuarial one.
How the ABS can claim an unemployment number of 823,000 for April when you have 1.6 million on the dole (not including those unemployed who cannot get the dole due to their partner working and those on the JobKeeper allowance – six million) is beyond me. Similarly, the ABS claims an increase of 118,000 unemployed for April as 600,000 jobs were lost.
Terry McCrann has attacked the ABS’ official unemployment figures, effectively calling them “fake” and claiming the real figure is more like 30 per cent in the private sector: “The ABS has to ditch its ludicrous methods of measuring joblessness if it wants to be taken seriously again” (Sunday Herald Sun, May 10, 2020).
He claimed that the ABS unemployment estimates were already ridiculous 30 years ago, and are beyond ridicule now. The only jobs data that now have any meaning come from Roy Morgan Research.
Back in 2017, Adam Creighton wrote in The Australian: “The definition of unemployment certainly doesn’t satisfy the ‘pub test’. It actually includes only a minority of people without work who want it. Imagine if a group of rogue statisticians, hellbent on issuing numbers that reflect reality, seized control of the Australian Bureau of Statistics. Their first decision would be to release an unemployment rate above 15 per cent — almost triple the official figure”.
There are many reasons for this but the following examples should suffice:
1: The ABS regard you as unemployed only if you are “actively looking for work” in the week prior to the survey period. But just what work is there to look for right now and indeed for many years, particularly since the global financial crisis? Not only are businesses not hiring now, many are being forced to not operate at all. Persons who only looked in newspapers or at job advertisements on the internet are seen as passively, rather than actively, looking for work and so are not considered unemployed. How job seekers are supposed to apply for a job that they are unqualified for or simply doesn’t exist, in order to be regarded as an “active” job seeker, is beyond me. Similarly, just checking noticeboards is not considered an active job search.
2: JobSeeker/Youth Training Allowance (dole) recipients, due to covid19 restrictions, do not at the moment have any mutual obligation requirements, (e.g. “actively looking for work”) because the requirements have been suspended until June 1, 2020. So, no one on JobSeeker allowance (1.6 million mid May) and no one on the JobKeeper allowance will be counted as unemployed. As of mid May, 835,000 employers, with six million employees, are covered by JobKeeper. The ANZ Job Vacancy survey for April shows that there were 64,000 vacancies, down from 136,000 in March and 62 per cent lower than a year ago.
Commentator Alan Kohler asked in The Australian on April 25: “Six million people are budgeted to get it (JobKeeper), which is 46 per cent of the workforce. Would they have been unemployed without it? Does that mean unemployment would have been 56 per cent, not 15 per cent, without the Job Keeper Allowance?”
3: If you work one hour in the week before the survey period you are regarded as being employed. The ABS gives the same status to people who work one hour as those who work 40 hours plus! A more objective measure would be the average number of hours worked per month per adult. Apart from being more objective, it incorporates the impact of unemployment, participation, underemployment and population ageing.
4: If you are not ready to start work during the week after the survey you are not counted as being unemployed. For example, if you have short-term health problems, are moving house or you cannot immediately obtain childcare, you don’t count. You are not unemployed.
5: People stood down who receive even a week or two of annual/long service leave will also be counted as employed. Additionally, people who have been laid off are not counted as unemployed if they believe they have a job to go back to within four weeks.
6: A falling rate of participation in the labour market. If people simply give up looking for work, that reduces the actual number of officially unemployed Australians. Half a million workers dropped out of the workforce in April and were not counted as unemployed. Hence the nonsensical ABS figure of 6.2 per cent for April.
7: Youth Allowance (youth dole) recipients who are studying part time as a requirement of receiving the dole are not counted as being unemployed.
8: If you receive JobSeeker (dole) but are allowed to engage in volunteering, work part time or are homeless, you are not counted as being unemployed.
9: If you have worked without pay in a family business during the survey week you are counted as being employed.
When you allow for these factors, the real figure for April was, well, certainly not 6.2 per cent. I believe it is at least around two million unemployed, or around 15 per cent, with 1.3 million underemployed. All up, around 25 per cent plus.
The dramatic understatement of Australia’s unemployment and underemployment figures by the ABS causes major distortions in handling covid19, economic planning and general policy making.
Certainly, Senator Kenneally wasn’t “dog whistling” when she recently called for a review of Australia’s temporary work visa system.
………………………………………………………………
All plausible based upon headline data that has been presented, but then lost me on the last sentence.
Where is the clear evidence on causation that (undefined) ‘temporary work visa system’ is to blame for stagnant income when couched within a whole array of other constraints e.g. lack of union coverage, nobbled award system, lack of compliance systems, bad management, shareholder demands for dividends etc.?
There has never been clear empirical support for the old trope that ‘immigrants take jobs’ when in fact immigration does support the economy, especially being ‘net financial contributors’.
It’s not about clear evidence of causation on one factor when all of those factors linked together cause the problem. If compliance was done, people (visa holders) wouldn’t be getting shafted in hourly rates, which also requires lack of unionisation, bad management etc. Your last sentence on immigrants being net financial contributors is true, but only up to a point. When they are being used as underpaid hands in retail or dogs body jobs they aren’t going to be net financial contributors. In fact there will be a statistical point, even in a perfect world, wher they are net financial contributors and then beyond that point a net financial drag.
This net financial contributors argument should not be swallowed whole, and should be considered against who is most harmed by current immigration, which is overwhelmingly the young and to a lesser extent (and who gives a toss?) the Executive Suite, where it is de rigeur to import a failed yank, Euro or other to our shores to stuff our universities and big corporates.
Do you have any specific, aprt from headline, data on these points that prove causation between (undefined) temporary ‘immigrants’ and the supposed negative outcomes.
Do you unerstand who ‘net financial contributors’ are, and how they support budgets?
Brilliant Marcus. Of most interest is that definition of ‘actively seeking employment’. If you’re doing that on the internet, you aren’t actively seeking, but how the hell do you find a job otherwise? Are you expected to walk the streets and walk into businesses and ask them if they have any work? That’s untenable, right there.
Almost as meaningless, is real estate data for PR presentation, masquerading as grounded and objective analysis.
Media are still largely using data from the first quarter while there seems only to be indirect data since.
For example, number of auctions, clearance rates, rental vacancies etc.; effectively suspended the ‘market’ when no valid nor reliable signals are available nor offered (vs. soem creative reporting on real estate as highlighted by Media Watch).
Why not class someone who thought about looking for a job in the last week as employed?