The latest – or some of it at least:
• It is now reckoned beyond doubt that the Liberals have held on in Chisholm, thereby guaranteeing a parliamentary majority of at least 76 seats out of 151. As related in the latest update in my late counting post, I think it more likely than not that they will supplement that with Macquarie and Bass, and wouldn’t write them off quite yet in Cowan. You are encouraged to use that thread to discuss the progress of the count, and to enjoy the reguarly updated results reporting facility while you’re about it.
• If you only read one thing about the collective failure of the opinion polls, make it Kevin Bonham’s comprehensive account. If you only read two, or don’t have quite that much time on your hands, a brief piece by Professor Brian Schmidt in The Guardian is worth a look.
• The three major polling companies have each acknowledged the issue in one way or another, far the most searching example of which is a piece in The Guardian by Peter Lewis of Essential Research. A statement released yesterday by Ipsos at least concedes there may be a problem with over-sampling of the politically engaged, but Monday’s offering by David Briggs of YouGov Galaxy in The Australian was defensive to a fault.
• Note the guest post below this one from Adrian Beaumont on tomorrow’s European Union elections in Britain.
1,724 comments on “Of swings and misses”
Roger Miller says:
Thursday, May 23, 2019 at 11:43 pm
No point being too fussy.
C@tmomma @ #1685 Thursday, May 23rd, 2019 – 8:48 pm
Not Sure @ #1684 Thursday, May 23rd, 2019 – 8:48 pm
Long time lurkers don’t throw their weight around and make demands of others when they decide to start commenting.
Forget the House, Labor will be a lost cause there until they can silence the hard left.
Albo will be a kinder more likable front. But the policies, well not AOC or Bernie Sanders stuff …. yet, but kinda still in the joke department some of them.
The fun is in the Senate now, as always.
Though I do expect there to be an investigation by the Senate into collusion between God and Morrison.
Plenty say the Canberra mob don’t listen.
With the problems that the pollsters have had for some time, no wonder they can’t read the mood.
On the few occasions I have been polled, I answered q’s as clear as I could.
Won’t be wasting my time being part of that again.
” Allowing Abbott to gain the upper hand by not calling a double dissolution over the greatest moral issue of our time. Weak options by Rudd which the opposition and MSM dined out on”…”
Swann and Gillard were vehmently against the idea in their little inner circle, which they leaked to the media, just to make sure Rudd didn’t change his mind.
I think in hindsight it is pretty clear Shorten & Co selected Rudd as the face most likely to beat Howard, but always planned to replace him, never expecting him to be so popular. As it turned out the stabbing of Rudd, having bought Labor out of the desert as it were, was again Labor rotting itself from within, and ironically the architect of it all got knifed in the front by the Australian people. If they had let Rudd do his terms, probably Gillard next we would still have a Labor govt. But the incessant undermining of Rudd pre his knifing was out of Shortens office.
Will Labor ever learn? Nope factional games and double dealing behind the scenes is more important than attaining govt… even against a hopeless Morrison govt.
So it will ever be.
Those deplorable uneducated Australians who don’t know whats good for them will still keep voting the wrong way.
Libs + L/NP + Nats + Cory + PHON = 38
ALP +Greens 35
Centre Alliance 2
Probably interesting to see Lambie and CA attitude to Coy tax cuts, energy costs and environment.
Does anyone know what happens in the senate if a vote is tied 38/38? Does the senate also have a speaker, who can cast a deciding ballot?
The President of the Senate votes normally (a “deliberative” vote), rather than having a casting vote as does the Speaker of the House.
OK you PB cyber experts. How do you prevent your comment disappearing while you go and google for some confirmatory or additional material to add to your brilliant post. This is driving me crazy. You work for as much as half an hour, go to check something, come back and the comment section is blank.
Adding to William’s comment, in the event of a tied vote the question before the Senate is negatived.
beguiledagain @ #1710 Friday, May 24th, 2019 – 4:13 am
I regularly copy my forays into the windmills of my mind.
I try to make it a habit although this is sure to fail when the item is reaching Nobel prize quality.
Thanks William and Itep!
Unfortunately your web browser may refresh the pager while you are not looking, and you lose everything.
The solution I have come up with is to, if I need more info before a post is submitted, get a text document up and post what I have in there.
I have occasionally lost lovingly hand-crafted posts which have taken ages to write, so I have learned the hard way.
But Kay Jay’s response is better 🙂
And now, I just can’t get that song out of my head!
On more psephological news from around the world, Narendra Modi of the BJP, has won a decisive (and surprise surprise) unexpected 2nd term in office today.
On reading this after work I realised why my very lovely Indian office mate looked so distraught this afternoon. I had noticed and wondered, but he obviously went through what I went through last Saturday. Unfortunately, none of these unexpected election results bodes well for scientists.
From the SMH:
Also, just to add information, my office mate has a Hindu name ( a reference to events around
Ghandi’s assignation long ago).
Modi wins big in India.
Apparently their polling is crap, too.
I have also spent a lot of time trying to understand the string of elections recently that have been “upsets” in terms of polling.
We are now looking at Brexit, Trump, Bolsanaro in Brazil, and now Australia.
There is also now good evidence from Avaaz that fake Facebook accounts are being viewed by enormous numbers of people for the EU elections.
I enjoyed Kevin Bonham’s article: http://kevinbonham.blogspot.com/2019/05/the-miracle-is-over-2019-australian.html
and agree with the conclusions.
My best guess about what is going on is that the “Cambridge Analytica” types are getting very good at targeting an undecided group of voters on social media, who are also swinging quite suddenly, and all in one direction, for a particular party / group.
The polls have not yet caught up.
It is not illegal, but it is driven by money from people with quite a lot of it, and who know how to spend it wisely to target the votes they need to switch. Also, anecdotally, I would say it is not just one demographic. I have collected a few friend and family stories about unusual messages / emails in the prepoll period, and particularly on Election Day.
I am also one of those people who make (some) of my living by delving into low signal to noise data, and using statistics to come up with defensible conclusions. After going back to the start of my field in circa 1970, I am really surprised at how right my discipline got it, using very little information, but including some modelling and sensible inferences, about what they were seeing. They had a few data points, but with many more datapoints we confirmed their results.
When you play with data a lot, you get an “feel” for what is possible and what is not. Sometimes collaborators send you an analysis, and you look at is and say “no this is wrong”. Sometimes it takes a while to convince them to take a second look, but that clanging sound in your head saying something is wrong is almost always correct. If you or they have stumbled on something new and wonderful, a few cursory checks will quickly tell you that the result is probably “Real”, and you go from there.
After talking to my colleagues on Australia over the last week, we all feel that something is “Wrong”. The opinion polls should have given us a better idea of what was happening.
William, Mark the Ballot, and Kevin Bonham noticed a problem with the poll herding in the last two weeks, and rightly urged caution about relying on the opinion polls.
I think opinion polls are extremely important for democracy. They are not gambling odds – they allow people fighting elections to read the public mood, and try to aim for the centre /consensus. Liberal democracy has relied, since WWII, on this centre consensus between parties to keep stability while advancing progressive causes, more slowly than many of us would like, but effectively.
Thursday, May 23, 2019 at 6:25 pm
Speers would be good for Q&A as well.
While I nearly always agree with your perceptive analysis of the shortcomings of our national public broadcaster, I have to disagree about Speers.
After watching him over the past five years, my impression is that he let the Coalition get away with too much, rarely testing them effectively on their fabrications and lies. He seemed to question Opposition spokespersons more sharply or amplify ill-founded criticism of Labor.
It’s only when we got closer to the election and it appeared that Labor had it in the bag that he started to challenge the Coalition.
I’m prepared to give him the benefit of the doubt because I assume like many in the Murdoch stable he felt he couldn’t stray too far from the party line. Now free from that yoke we might finally see the real David Speers.
My choice for either post would have been Karvelas, with perhaps Ellen Fanning or Emma Alberici for Q and A. I am glad, however that that supercilious duo Rowland and Trioli will not be the running.
You have to admit that David Speers did a magnificent job on George Brandis, back in the day.
Friday, May 24, 2019 at 7:00 am
You have to admit that David Speers did a magnificent job on George Brandis, back in the day.
I remember viewing that interview at the time. I disliked Brandis as much as anyone and was annoyed at the creeping intrusion of the state into our lives in the guise of national security. But I was surprised that it was considered a “train wreck” and that it won a Walkley.
I’ve just viewed it again. While I’ve had a P.C. since 1981 and my 12 year old son urged me then to buy a modem, years before the internet became a fact, I still only see it as confusion between Speers and Brandis on Metadata, a very complex subject that most people hadn’t any idea of. Perhaps I was missing something.
Brandis was trying to dispel the idea the government was spying on the content of your communications and I think he did that fairly well, although awkwardly at times.
Speers got on this kick of differentiating between websites and web addresses (IP) and trying to suggest that the security agencies would be able to find out where you have been surfing. I thought Brandis was agreeing with that while insisting strongly that the spooks weren’t interested in the content of what you were watching, which was something that surprised me in a terrorism context.
As a former journalist, I’m also disturbed by the trend in journalism awards to recognize “gotchas” or “train wrecks.” I’m all for journalists being persistent, but that kind of approach generates more heat than light. I think this is best illustrated by comparing the interview styles of Kerry O’Brien and Leigh Sales. I think you can figure out which one I prefer.
Still, I’m hoping Speers at the ABC will be able to fly free from the constraints of working for an agency of Rupert Murdoch.
OK you professional psephologists, have I got this right. I was never much good at mathematics:
Take a look at the numbers from the latest count.
The hayseeds and bogans won the election for the Coalition.
Coalition (all but Qld and WA) 44 (includes 10 hayseeds and bogans in NSW and Victoria)
Labor (all but Qld and WA) 56
(Qld and WA)
Coalition 34 (23 of them in Queensland)
Labor 11 (mainly urban seats)
Could someone come up with the urban-rural split before the election, and the same number after the election.
Is it likely that Labor will be able to snag some of those 34 seats? And is the coalition’s 44 seat total outside of Queensland and WA the lowest they will go?