The only new poll this week was the regular result from Essential Research, but it was enough to contribute to another sizeable cut in the Coalition two-party lead for the fifth week in a row. Sharp-eyed observers will note that the state seat tallies now account for redistribution changes, which have added a seat in Western Australia and removed one in New South Wales. These changes have seen the abolition of a Labor-held seat in the Hunter region of New South Wales, and the creation of the notionally Liberal seat of Burt in Perth. However, the overall effect is favourable to Labor since three seats in New South Wales – Barton, Paterson and Dobell – have become notionally Labor on the new boundaries, with respective margins of 5.2%, 1.3% and 0.4%. The swing currently being in Labor’s favour, the model rates them a certainty in Barton and better than evens in Paterson and Dobell, and more likely than not to win to win Burt.
The upshot of all this is that BludgerTrack has the Coalition down three seats this week in New South Wales and steady everywhere else, whereas Labor is credited with two gains in New South Wales and one in Western Australia. Note that the national and state-level figures on the chart showing seat change since 2013 will no longer align, since the baseline for the national result is as per the election (Coalition 90, Labor 55), whereas those for the state numbers are post-redistribution (Coalition 27, Labor 20 in New South Wales; Coalition 13, Labor 3 in Western Australia). The post-redistribution margins are as determined by myself, following very similar methodology to Antony Green. A full accounting of the calculations can be found here for New South Wales, and here for Western Australia.
Other news:
• A ReachTEL poll of 712 respondents in New England, “obtained by Guardian Australia” (who commissioned it is not clear), suggests Barnaby Joyce would have a very serious fight on his hands if former member Tony Windsor sought to run again as an independent, which he is neithe ruling in or out. The numbers cited are 39.5% for Joyce, 32.2% for Windsor, 11.2% for Labor and 4.6% for the Greens, with 5.1% undecided.
• The mass exodus of Labor’s Western Australian federal MPs continued this week, with Senator Joe Bullock announcing his decision to retire in protest over the party’s support for same-sex marriage. Bill Shorten promptly announced that Bullock’s vacancy would be filled by Pat Dodson, a leader of the Yawuru people from Broome and former chair of the Council for Aboriginal Reconciliation. This scotched the ambitions of Louise Pratt, who was famously relegated to second position on the Labor ticket behind Bullock, then defeated at the April 2014 Senate election re-run following a collapse in Labor support. Many attributed this outcome to derogatory comments Bullock made about Pratt while speaking at a Christian function, which became public the day before the election.
• Labor has another indigenous parliamentarian lined up in the form of Linda Burney, who has held the seat of Canterbury in the New South Wales state parliament in 2003, and served as Deputy Opposition Leader since the defeat of 2011. Burney is running for preselection in Barton, which encompasses about half of her current electorate. The seat is currently held for the Liberals by Nick Varvaris, but Labor has been heavily favoured in the redistribution, which adds inner city territory around Marrickville and removes Liberal-voting Sans Souci. Burney has resigned as member for Canterbury to contest the preselection, which will result in a by-election.
• Mal Brough’s announcement that he will not seek another term has opened a Liberal National Party vacancy in his Sunshine Coast seat of Fisher. The party’s state executive had been withholding endorsement of Brough’s preselection pending the outcome of an Australian Federal Police investigation into his role in the leaking the diary of Peter Slipper, the then Speaker and his predecessor as member for Fisher. It was promptly suggested that Jarrod Bleijie, the controversial Newman government Attorney-General and member for the local electorate of Kawana, might be interested in the seat, but he has since ruled himself out. Amy Remeikis of Fairfax reports the seat might be of interest to James McGrath, who ran against Brough for the preselection in 2013 and has since found a place in the Senate.
On Turnbull’s visit to the Islamic Council – firstly it is a good thing that Turnbull has, if nothing else, changed the tone in that area. Abbott was destructive in many areas of Australian life, but none more so than relations with Muslim Australians, so for that I’m thankful to Turnbull.
However:
Surely by now it is clear that if Turnbull is coming across as “humble” he is automatically not being “genuine”.
Nice work, yabba
Nikki Savva is standing up very well to Davis Speers’ questioning about her book.
Hi Yabba – just wondering out of interest, if you might mention a couple of the software tools you use for modelling?
[sorceror43 @sorceror43 9m9 minutes ago
sorceror43 Retweeted Peter van Onselen
No Newspoll this week, @mumbletwits ]
Next week also confirmed by PvO.
Includes a neat little video on the construction of skyrail.
[In a surprising show of solidarity, the RACV, Bus Association Victoria, Bicycle Network, Victoria Walks and the Public Transport Users Association have all endorsed the Andrews government’s elevated rail plan to remove nine of Melbourne’s worst level crossings.
The plan has provoked a fierce backlash from residents, who fear the rail viaduct will build a barrier between communities, attract crime and graffiti, increase noise from trains and reduce property values.]
The Opposition are going to try to kill it for political purposes, of course.
http://www.theage.com.au/victoria/sky-rail-on-a-high-as-transport-groups-back-elevated-rail-plan-20160307-gnccpt.html
Modelling is a powerful tool to analyse loads of data, processing it according to programmed rules. It can be used to test the effects of changes to the rules or to the inputs. Depeding upon the application, the rules might be the laws of physics or they may be assumptions about how the economy (or groups of people) behave.
Like any computing application, it’s only as good as its programming and data. A general statement about modelling, as with any other computing application: garbage in, garbage out.
Confessions @2655, I’ll be back in a minute, I’ll grab my tinfoil hat from the stand in the hallway….
Savva and her lawyers are too smart to say Abbott and Credlin had an affair. They just leave it as a conclusion the reader may draw as well as saying there were politically damaging rumors of an affair which I’m sure they can back up.
Labor lost that debate when they allowed a “rail Bridge” to be called a Skyrail, as if it is something out of the ordinary.
Humans build many bridges. Sometimes cars go over them, sometimes trains do. It’s not a big deal. Calling it a Skyrail just makes it sound more unusual, and thus gives people the ability to protest.
Dio@2659: “Savva and her lawyers are too smart to say Abbott and Credlin had an affair. They just leave it as a conclusion the reader may draw as well as saying there were politically damaging rumors of an affair which I’m sure they can back up.”
Something like that. They remember what happened to Bob Ellis. As should we all.
guytaur
Hydro is renewable – yes. It’s been a great option since the year dot but it doesn’t meet Tassies needs. If they had more renewables they could keep hydro in reserve.
Libertarian Unionist
[Tassie gets the bulk of it’s energy from hydro and imports.
PV and wind were only ever going to supplement this across the year, particularly as the wind resources need additional network expenditure to be integrated.]
PV and wind can deliver up to 100% of SA’s energy – typically quoted as 50% on average.
Better to invest in additional infrastructure rather than pissing money against the wall buying fossil fuel. Battery storage is much less of a requirement when you’ve got hydro.
Hydro can be used to store excess energy generated by renewables by pumping water uphill. The Snowy Mountains scheme has been doing so since it was built, buying energy when it was cheap and selling when it was more expensive.
[During the Rudd-Gillard years, Savva recalls: “I was the darling of the conservatives; they all thought I was wonderful because I was getting stuck into the Labor Party.
“Abbott would text reasonably frequently to say, ‘Good column, very shrewd observations’.”
But as her columns in The Australian became increasingly critical, she says she transformed into the “she-devil” of the Abbott government.
“Even those of us who thought we were insiders learnt something reading her column,” says one of the key players in the Turnbull coup. ]
http://www.theage.com.au/federal-politics/political-news/niki-savva-how-journalist-went-from-conservative-darling-to-abbott-she-devil-20160307-gnc6vl.html
Can anyone tell me what this “pizza in a cup” is in relation to #ideasboom?
Meher
Does Bob Ellis also say conservos don’t engage in infidelity.?
Did he get sued because some conservos took offence at his claims that they were asexual!
I guess Mr Profumo would have been the first to sue had he been alive and Australian.
Maybe Mr Pell’s inability to see what his brother priests were up to was genuine, because of his conservo naivety.
Because you have today educated me, I now know that Jim Cairns, a leftie, was certainly shagging Ms Morosi, but John Gorton, a conservo, was not shagging Ms Gotto.
There is a story about Snowy Hydro taking a busload of visiting engineers to have a look a the dams and the potential – sometime in the early 60s.
They were all blokes of course.
Naturally, being engineers, they all hit the turps which necessitated the bus stopping here and there.
One of the visitors was reported to have been a bit reluctant.
When challenged about his shyness he replied, ‘Because you bastards will probably dam it before I have the chance to finish shaking it.’
shiftaling 2654
For big financial models, I used to use Mondelio. Now use my own comprehensive system which sits on SQLS. For time lapse simulation and discrete event simulation I also use my own proprietary frameworks which sit on SQLS, and are mostly programmed in TSQL, with Excel purely as a front end and reporting tool. There is a bunch of VBA which handles Excel user interfaces for clients who require a familiar face.
For LP and IP I use my own re-creation and embellishment of an ancient super-smart model building tool called OMNI, which creates MPS files which can be input to any LP/IP Solver program. For LP and IP solving I use LPSolve, but there are many available. Simulated Annealing ia another matter, which I will not go into.
cheers,
Scott, Skyrail is hardly a mere bridge. It’ll be more of a why-a-duck. And nothing wrong with that – there will be long openings under it, between the whale-tail supports, so it will in fact “divide the suburbs” much less than a ground-level rail line.
Re modelling: I won’t try to answer all the comments.
Yes, of course modelling is only as good as its inputs and assumptions. Econometric modelling is an inexact science, because there is a lot of uncertainty about the relative significance of different inputs and different assumptions.
And the same largely applies to climate modelling: a lot of the climate modelling so far has been way out, for example in relation to the rate of acidification in the oceans, which has been significantly overestimated so far. But that’s ok, scientists learn from discrepancies between modelling and reality, as do economists.
Any modeller who claims that they are producing an accurate prediction of the future is an idiot. JK Galbraith rightly once said that the only point of economic forecasting was to make astrologers look good.
But modelling can play a useful role in assessing the possible impact of policy changes in areas such as tax. There are two aspects of such modelling
1) the potential impact on revenue and the distribution of the tax burden. Some credible work has been done on these in relation to the ALP policy both by the McKell Institute (as one scenario among a number they considered) and the ANU (specifically in relation to the policy announced by Shorten and Bowen). But neither of these exercises looked at all at:
2) the potential impact on the housing market. That is, all things (especially interest rates) being equal, what the policy might do to the investment/purchase choices made by Australians and thereby house prices and market behaviour.
There has been a lot of modelling work done on the Australian housing market over the years, so there would be no difficulty with running these tax changes through the model to see what comes out. What we would see wouldn’t be any sort of exact prediction of the future, but it would tell us something about whether or not there might be upward pressure on rents and downward pressure on prices for existing dwellings.
And, no Briefly, no such modelling of the impacts of neg gearing changes has been done for a long time. Except perhaps by BIS Shrapnel who, despite all the dumping on them that has taken place in the past few days, are reputable modellers. But, although I haven’t seen it, I don’t think their modelling is going to be much help, as it doesn’t include any assumptions about CGT changes and I understand it doesn’t look at the question of neg gearing being restricted to new dwellings. Morrison’s use of it was therefore a beatup and he deserved to be criticised.
While I’m at it, I also don’t agree at all that negative gearing has caused the current “problems” with the housing market in Sydney and Melbourne (there are no signs of any such problems where I live in Tasmania and, hey, we are part of Australia too). In my view, these are mainly the cumulative result of a number of factors: rising real disposable incomes among the middle to upper income households, persistently record low interest rates for a number of years, high levels of competition among lenders and, of course, an upsurge in the amount of foreign money coming into the market.
It could be argued that removing negative gearing might be one way of taking fuel off the fire. But, in that context, state governments significantly increasing land tax would work equally well and would arguably be more equitable.
Anyway, as I have said a number of times, my main concern is that I have no certainty as to what effect Labor’s policy would have on the housing market and neither do they. It’s an experiment.
And, as I said in an earlier post, economic experiments can be much more dangerous than scientific ones. Ask the populations of Soviet Russia or of Chile under Pinochet, both of whom were subjected to a whole raft of mad scientist economic theories (the Russians to left-wing theories and the Chileans to the more extreme forms of economic rationalism).
BTW, to be clear, I used to have a negatively geared property but no longer. My views are not in any way based on self interest but on a concern for well thought-through, evidence-based economic policy. Which, I’m sorry to say, we don’t have here.
Dio
Savva needs to say zero about the Abbott and Credlin relationship. Connie F-W gave it all to her on a plate, by simply telling her of the conversations she had with each of them.
[meher baba 2584
Scientific experimentation is very different to, and far more accurate than, economic experimentation. But modelling is modelling: the only accurate test of it is how it compares to the reality that eventually comes along.]
The best scientific theories (QED and relativity) can predict experimental/observational outcomes to about 10 decimal places.
Economic modelling can’t reliably predict outcomes to even a single decimal place, and probably never will.
Chalk and cheese.
Thanks Yabba
Millenial @ 2644,
Could you please put Fiona Scott in Lindsay!?! One Lucy Wicks is more than enough! 😀
Sounds like a lot of black boxes!
New thread.
What about Ipsos Fairfax?
Re Tony Rabbott and Warringah – a friend has reminded me that at any time a person (he thinks a voter in Warringah but I don’t see any “standing” requirement) can sue for $200 under the Common Informers (Parliamentary Disquals) Act 1975. See http://www.austlii.edu.au/au/legis/cth/consol_act/cida1975507/s3.html Go for it, someone! But as said friend says, you’d have to be ready to withdraw and pay a few solicitor bills if/when Abbott provides proof of renunciation or letter from the Brits.
Steve777
Real modelling is not data analysis. Properly done, modelling recreates the actual processes going on in the real world in sufficient detail that the modelled outcomes sre very close to actual outcomes of the system being modelled. Physics rarely comes into it, unless you are modelling physical systems, like car suspensions, hydro power generation or electric power distribution networks, all of which I have done.
Commonly, useful models are based on flows, stocks and proportions which vary through time, and which are able to be related by known, or easily defined formulae. Feedback loops and leads and lags are often involved, and getting these right is very important. Flows and stocks can reflect quantities of physical things, like numbers of people, or kg of protein, or loaves of bread, or cubic metres of concrete. (All per unit of time, for flows). They can also be money. Proportions are things like prices, interest rates, pay rates, overtime loading percent, protein concentration in kg per 1000 litres of milk, loaves per dough, flour recovery per tonne of grist, cement kg per cubic metre of 20 MPA concrete, and so on, and on and on.
Useful models are often very large, but they follow basic, strictly logical rules. Building them is a combination of art, intelligence, mathematics and lots of patience. Experience is particularly useful. The discipline is called Operations Research.
Meher
You say that Labor’s policy to reduce negative gearing concerns you because it is an “experiment”.
So have we reached the stage that before we even start to reduce the generous gift that negative gearing is to investors, we must prove to some high degree (almost beyond reasonable doubt) that there will not be some imaginary fatal effect on the market.
Maybe this is a case where the lesser of two evils should be chosen, and common sense should be the guide.
So what if house prices drop 5 or 10% across the board. This will all smooth out for homeowners, and they’ll take it in their stride.
Only the speculators, those seeking quick capital gains, the “day traders” of the housing market will have to adjust their sums.
Have you never sat in a pub, or with a group of friends, and heard their discussions about how great negative gearing is, because they now “own” 2 or 3 or 25 houses which the renter and the government are paying off for them.
This is the starting point in deciding if negative gearing should be modified. Should first home buyers who have to pay off their mortgage with after-tax dollars have to compete with those who have a government assisted greater capacity to pay off a higher mortgages.
One thing is for sure in the minds of all of us who don’t know so much about modelling, (except that garbage in gives garbage out) . That is the current system is unfair to young couples buying a first home. If others suffer a bit because of changes, and these youngsters nenefit, then good-oh.
Last line is “benefit”
meher baba @2460
I have not read what Ergas wrote most recently, but have read much of what he has written in the past. The man is a skillful liar of whom people should be very wary.
The Australian Competition Tribunal (an entity associated with the Federal Court and presided over by a Federal Court judge) found that he has “an inability to express an objective expert opinion upon which reliance can be placed”.
It has been said that “Everyone is entitled to his own opinion, but not his own facts”. The particular skill of Ergas is to disguise his own facts within his opinions, thereby elevating the opinion to the level of (supposed) fact. For example in a relatively recent piece he (without explanation) substituted gross national income (GNI) into a context where gross domestic product (GDP) is used universally and then based his argument on the results of this substitution (of GNI as “his own fact”). And this is not an isolated example; in every piece I have read from him he has introduced unusual or irrelevant facts, used cherry-picked data and otherwise engaged in scientific fraud (economics claims to be some dismal sort of science, I believe).
meha @ 2584
[ratsak@2569: “As someone noted a very different thing to scientific modelling where the model method and all inputs are available for peer review.”
No, credible economic modelling is identical to this sort of modelling. And it is equally subject to peer review.
Scientific experimentation is very different to, and far more accurate than, economic experimentation. But modelling is modelling: the only accurate test of it is how it compares to the reality that eventually comes along.]
Science is science; economics-whilst claiming to be a science-does not require its practitioners to behave like scientists, and they do not.
For example, many current economics textbooks contain an incorrect theory about the process by which commercial banks create money (as does the introductory course at the Harvard Business School). In particular the incorrect theory (money multiplier) has the causality running in the opposite direction, and this is contradicted by evidence.
This would be equivalent to physics continuing to teach phlogiston, aether or the approach of Aristotle, which does not occur.
It is not equivalent to continuing to teach Newtonian mechanics, which is merely inaccurate rather than wrong. Newtonian mechanics has considerable predictive power and is thus extremely useful except in the situations where its limitations are exposed. Theories that get the causality wrong have no predictive power (rather the opposite).
So the question is why does economics continue to teach things that economists to be wrong becuase they are contradicted by evidence? One answer – it’s easier for students to understand – doesn’t wash in physics. Another answer – it would (in this particular instance) offend powerful and wealthy people who learned it at Harvard Business School – is closer to the mark, and that shames the whole discipline: peer review is irrelevant if enough of the peers are corrupt.