Employment thumps the bullhawks

The total number of people employed in Australia fell by 22,119 in April.
This is the third time in the past 5 months we have seen the total number of people employed in Australia fall. Yes, that’s right, the 3rd time in 5 months.
The market was looking for a gain of 17,000 jobs. The fall of 22,119 was driven by a 49,060 in full time jobs while part time rose by 26,941.  However there was still real job losses in excess of 22,000

Sadly, we can’t do a sectoral breakdown until next month but for now there’s no real surpise that the weakness is clearly in the services economy states of NSW (down 22,000) and VIC (down 17,000). The mining states held up better: QLD (down 4,000), SA (up 1,000) and WA (down 4,000).

Importantly after adding over 428,000 jobs in the 12 months to November 2010, jobs growth over the past 5 months has totaled 25,500.


Hours worked were also down over the month.  
God knows what the bullhawks (Joye, Bloxham, Carr) will make of this data but it clearly is not an economy headed into a capacity contrained boom. It is an economy in adjustment. There’s no saying that that will be smooth, nor that using big macro tools like interest rates and the currnecy won’t result in big overshoots, most especially when trying to juggle that huge household debt.

You can forget a June rise on this data.


    • Armand Tamzarian

      Yeah this bloke was straight out on Business Content Mill with a self serving justification of his “analysis.”

      Message to Adam Carr: just admit you have no clue.

  1. Carr:

    Look, in this environment, when plenty of people are out there trying to jawbone the RBA and prevent further rate hikes – telling you that retailers are suffering miserably and that the economy is weak and fragile – this data will inevitably be misused as ammunition for that case. In my opinion though, that is a fairly green error.


    • “the dovish case is simply not credible,
      under any reasonable assessment”……..cheers JR

      • Armand Tamzarian

        One of the (many) reasons that website has little credibility (IMO) is their persistence with that guy as content contributor (as opposed to analyst).

    • Deus Forex Machina

      He really said that?

      What a self righteous condescending know it all…wonder if he apologised to the guys he would have had short interest rates running into the number and got their backsides handed to them…

  2. If the economy is on two speeds we seem to need policies that don’t apply to the whole economy (like interest rates) but that are able to differentiate.

    • A mining tax would have worked fine but the government got greedy and stupid.
      Not that its the miners fault we have a property bubble where the debt has finally come home to roost.

  3. A.Carr has no shame. He should be embarrassed.

    But the 0.8% fall in participation in NSW looks to be an ABS statistical error

    Meaning the unemployment rate should’ve been around 5.1% instead of 4.9%

    • Deus Forex Machina

      not sure…as someone who lives in the hunter i’ve noticed that 5 of the last 6 months there has been a fall in employment…

      the next batch of localised data is out next week…which will be interesting

      but if, as it is, newcastle is a microcosm of the australian economy with a kicker from mining in the valley…what the hell is going on

      i reckon its more than statistical noise which is what all the major bank economists and blxham have said so far

        • No the hard numbers are quite possibly a trend. Hard to tell from one month though – next few months will decide that.

          I’m saying the participation rate is an error…there is no way it can change so much in 1 month. It was either too high before and has come back to what it should be or its too low now and will be fixed next month.

          • Deus Forex Machina

            BK…see comment below i posted for cameron…was meant to be here…sorry

          • Why not?

            Other illiquid stats can change “significantly” in a short period of time (eg. housing listing stats, of which i collect a lot). I’ve frequently seen, for example, weekly rental spikes of ~3% (and even some sales listings spikes) in the vicinity of “stress events”.

            Could it simply be that this “noise” or “error” is legit data, gathered around a systemic “stress event(s)”.

            You may well be correct, but i thought i would put another thought out there for the “big, illiquid systems that suddenly change quickly” POV.


      • the ‘trend’ that you see may be actual and this information may have a darker side as when you reflect on the previous boom/busts in other countries, house prices peaked in the USA before employment started to fall (ref?) but here in different Australia, we have a significant portion of people attached to housing. If you read the recent Nab survey for the march qtr the info seemd to say, retail, construction, transport and real estate all had a bad qtr. This would indicate to me that there could have been a fall in employment in many sectors of the economy.
        FWIW, I had a part timer leave and replaced with a full timer (who came from a real estate office) and last week admitted that the wages were paid in the last few months intimitantly and only after a sale had been made, once she went 3 weeks without being paid !! and her holiday exit pay is being paid out over 12 weeks as there was no cash to pay her the 4 weeks pay she was owed….she jumped before the cash dried up completely.

  4. FWIW, these are the seasonally adjusted numbers for the past 9 months. (Revisions after the –>).

    Things to note:

    – huge revision down in January.
    – employment was flat between Nov 10 and Feb 11
    – March got revised up a tad, so it makes this months fall seem bigger

    I suspect April will be revised up as well next month.

    Apr 2011: 11 436.5 –> ?
    Mar 2011: 11,457.1 –> 11 458.6
    Feb 2011: 11,412.9 –> 11,419.3
    Jan 2011: 11,441.5 –> 11,423.0
    Dec 2010: 11,417.4 –> 11,417.5
    Nov 2010: 11,416.9 –> 11,415.1
    Oct 2010: 11,355.9 –> 11,362.3
    Sep 2010: 11,326.2 –> 11 324.9
    Aug 2010: 11,272.0 –> 11,275.4

  5. Chris Joye says:

    The truth is that the monthly UE data are a crap-shoot with a wide 95% confidence interval for total jobs created of between -77k and +32.5k (the actual number was -22k)

    Which is exactly the point I made last month, when the bullhawks were cheering the 37,800 extra jobs.

    The (unrevised) data is very noisy, and the ABS says with a 95% confidence interval the result for March (+37,800) could be anywhere from -16,800 to +92,400. Hilarious really given how much the media carries on about the monthly ups and downs.

    • Armand Tamzarian

      It is even more hilarious that blokes like Joye claim to recognize the uncertainty in the numbers but speak with such certainty (and arrogance) about there being full employment and therefore (along with other misinterpreted data) a need for rate rise.

    • Great points Lorax. That is why many of the authors and commentors here like to looke for trends, but of course once the trend is established in the data, it is too late to act on it.

      And you make an important observation that it is the side whose views are not supported by the headline figure usually bring up the statistical argument. Although I don’t remember it happening last month.

      My views on such matters are here

      • Deus Forex Machina

        Guys…light blue line top chart…thats the trend of the monthly change in employment and should remove the old excuse of “monthly volatility”

        don’t let the thin slicers tell you this is just noise because they missed it…it is true that this is a very volatile data release with a standard deviation above 20k…

        I stopped trading this number in 1990 or 1991 because forecasters were so hopeless at it and I kept getting my backside handed to me

        but the forecasters have the same data that sword has…its abs data the light blue line is still the light blue line and all they are doing is justifying why they were wrong and why they are still right on rba hikes

        but thats what they say isnt it…lies, damned lies and statistics 🙂

  6. More reasons why it cannot possibly be so here.

    Look, they may well be right, it might be an outlier, but their reaction speaks volumes about their ability to independently analyse economic data.

    • Armand Tamzarian

      I couldn’t get past “Variously described by News Ltd as a Svengali and one pollie’s economist muse” without being physically sick.

      It is amazing how analytical failure can bring success in economics. If engineers build bridges that fall down they don’t get re-employed. In economics that sort of record seems to actually enhance your reputation.

    • Yeah its funny Armand.

      You see in trading, getting it right more than 50% of the time is luck, whereas in engineering and economics it would be negligent.

      BTW, that same muse is trying to trademark “The Great Inflation” phrase because he reckons we are about to head on into a great phase of inflation due to wage pressures and a strong economy.

      Even though it was used 50 years ago. And in the 1920’s.

      Oh well, it takes his mind off house prices….

    • Yeah mining is experencing a shortage and the rest of the economy is shedding jobs. Did everyone read about China’s imports of commidities is start to slow down. That, job loses and any other interest rate rises will crash the housing market even faster.

      I do recruitment in the mining industry and spoke to one of the big companies yesterday they are looking everywhere to find people all over the world. Then you read this and it really comes home to you.

      • Armand Tamzarian

        Even allowing for seasonal trends China’s copper imports have been down for awhile on a rolling YOY basis — you won’t read much about this in bullish stuff about copper only going up from here because of supply shortfalls. I’ve been reading about copper deficits for more than a year now but all I’m seeing is stockpiles trending north. Go figure.