December 2 Provisional vote changes in 2007

As you know, the Howard government got tough on provisional votes at the 2007 election. As you also know, they did it for party political reasons. But did it make a difference? 

Recall this from soon after the election, in which I suggested that the rule changes cost Labor 'a few' seats.

I since did some calculations, and was surprised to see that it probably didn't make much difference. Most likely just one seat, in fact. 

I made these assumptions in each seat: that if it weren't for Proof of ID and the change in counting rules, the number of provisional votes cast would have increased by ten percent on the 2004 election (as they had roughly increased from 2001 to 2004); that the same proportion would have been counted in 2007 as were counted in 2004; and that in 2007 the missing votes would have gone, in two party preferred terms, the same way as those that were counted (in 2007).

It made little difference as table below shows. Column on right is my estimate of ALP two party preferred making above assumptions. Only McEwen changes hands.

2007 close seat results with adjusted provisional votes

Note McEwen has a new result after the court case, which increased the Coalition's lead. I don't have the new breakdown of how many were provisional votes, and so have not put the new figures in.

The Federal Court's decision may have may it zero instead of one. McEwen needs more research.

Nationally, Labor's vote goes from 52.70 to 52.75, which in uniform terms would yield two extra seats. Labor's advantage in provisional votes seems under-represented in the marginal seats.

(Note that in La Trobe the ALP goes backwards on my calculations. It was unusual in the Liberal doing (a little) better in provisional votes than in all other votes.)

[Update: In some seats - a little over half of them in Qld - the number of provisional votes cast did increase more than ten percent. That's a flaw in the formula; I just tried using actual 2007 votes received plus 20 percent, but even that yields only McEwen. (Pushes Bowman up to 49.986.)]

Am willing to be corrected on all this. 

November 27 Those *&*@ Howard Years (again)

Me in Inside Story, kind of an elaboration on earlier Crikey piece (and taking in Episode 2).

Here. [December 1 update: also in the Canberra Times.]

November 26 2008 My bets: Victoria in 2010

Yesterday I took a bet from a Victorian, who gave three to one against my expectations expressed in this recent post that Ted Baillieu would lose the Victorian Liberal leadership before the next state election, but that his party would win that election. (So both things have to happen for me to win.)

This person obviously reads his papers and has more knowledge of Victorian current affairs than me, but as you know I am a 'pattern guy', preferring forests to  trees. (And not averse to metaphor-mixing.)

I am less sure of the second outcome than the first, but I just don't think opposition leaders last very long these days, and expect both Baillieu and Malcolm Turnbull to be gone before their next elections, probably by the end of next year. (In Malcolm's case the proviso, as always, is no early election.)

NSW is different: no sane person would back Labor to win in 2011. But I also worry about Mr O'Barrel, whose challenge is to stay in the job for another 28 months. If he does that, he'll be Premier.

People always say 'oh, the opposition shouldn't/can't just wait for the government to fall over, they've got to present a viable alternative'. But generally oppositions don't really do that, and most changes of government are of the 'drovers dog' variety.

November 24 Newspoll says 55 to 45 (yet again)

In the Oz; George Megalogenis here. Tables at Newspoll here.

[Update: Here from Newspoll is the budget deficit question.

"THINKING NOW ABOUT THE AUSTRALIAN ECONOMY, AND IN PARTICULAR THE RECENT FINANCIAL CRISIS. THE FEDERAL GOVERNMENT PLANS TO INJECT ADDITIONAL SPENDING INTO THE ECONOMY NEXT MONTH TO STIMULATE ECONOMIC GROWTH, WHILE STILL MAINTAINING A SMALL BUDGET SURPLUS. WOULD YOU BE CONCERNED OR NOT CONCERNED IF THE FEDERAL BUDGET WERE TO GO INTO DEFICIT AS A RESULT OF ANY FURTHER GOVERNMENT SPENDING IN THE NEW YEAR? "

Wording such questions is always tricky. For example, putting a 'small' before 'deficit', as there was before 'surplus', would probably have lessened the amount of concern.

Economics 101: next year's budget deficit

' Economics 101' being the level I operate on, but ... everyone seems to be ignoring one fact:

One-off spending this financial year has zero effect on next year's budget deficit, doesn't it? (Apart from, around the margins, interest on investments or whatever.) So next month's $10.4bn splurge has nothing to do with the 2008-9 deficit, and nor would further spending early next year.

And perhaps the polling question could have included something like 'this spending will not effect the 2008-09 budget outcome at all'.

(Am willing to be corrected.)

November 22 First anniversary: in this month in 1997

That is, one year into the Howard government.

Three ministers had resigned, the most recent being Bob Woods. Mal Colston and 'work for the dole' were in the news. 

The words

Paul Kelly in the Oz:

"During the 1980s there was an orthodoxy about Howard - that he was the reforming spearhead of the Coalition, but weak on politics. Yet the orthodoxy today is more likely to be the reverse that Howard is an adept politician with doubts about his ability to implement genuine change."

In the SMH, Geoff Kitney had a soft interview with the PM, of which perhaps the most interesting line is:

"[Howard] is learning, mainly by the mistakes he and some of his ministers have made, that international policy is vitally important to Australia's future."

And over at the Age, Shaun Carney on the then opposition's prospects: 

"On any reading of political history, the Labor Party can expect to be consigned to Opposition for at least two terms ... [They can either work hard or take] the course that John Howard took. That was to convert most policies into platitudes and let the incumbent Government slowly starve itself of energy and oxygen. But that took 13 years."

In the end it only took 11 and a half.

The opinion polls

The most recent Newspoll, late Feb early March, had the Howard government way ahead 49 to 37, which two party preferred would have been about 55 to 45. 

That's the same as Labor's lead these days. On 'satisfaction', Howard was on 51 to Beazley's 46, as against 65 for Rudd and 53 for Turnbull in the most recent 2008 Newspoll. On better PM, Howard led Beazley 53 to 23; Rudd leads Turnbull 62 to 22.

(These are only individual poll numbers of course. The 1997 numbers represented a slight jump in the government's fortunes, while polls today are rather flat.)

However, after March 1997 the polls moved Labor's way, in May hitting 50-50 before dipping down again, but in October (after Howard's GST announcement) Labor took the lead and Beazley was (just) preferred to Howard as PM. 

The ALP pretty well held the lead up to the 1998 election (which it won in votes 51 to 49).

The Newspoll graphs

As this graph shows. (My two party preferreds from Newspoll primaries; the first and last pairs of numbers represent election results.)

And below is Newspoll over Rudd's first year. (Newspoll's 2pps this time; first numbers are election result.)

Malcolm Turnbull will be hoping for a similar trajectory from here.

November 21 Nielsen in Victoria: 55 to 45

In the Age here, table here.

Recall the most recent Newpoll, which had had 51 to 49.

I reckon the following two things are more likely to happen than not:

- Ted Baillieu to lose the Liberal leadership, and 
- (more contentiously) Labor to lose the next Victorian election.

But we have to wait two very long years to find out.

November 18 pm 'Howard Years': me in Crikey today

On the weird theme of the first episode, here.

(If you have a Crikey login, it's on their site, with readers' comments - mainly on the show - here.) [Next day's reader comments in newsletter here (scroll down a bit).]

November 17 Nielsen and Galaxy say 55 to 45

Links - from Pollbludger - here and here.

'The Howard years' on ABC TV

First episode tonight. The ads show a bunch of politicians, 'on message', doing what politicians do best: espouse clichés for journalists to build stories around. The title indicates that, despite last year's election, the Howard cult of personality remains. 

1993's 'Labor in power', mainly filmed (I think) during the term 1990-93 when most in the government believed they were headed for defeat, at least had the Hawke-Keating split to generate some spark and dissent.

And why are George Bush and Tony Blair even there? As if they'll be anything other than diplomatic.

Still, that's just from the ads. Shouldn't prejudge!

NSW, the 'sophomore surge' and the 2010 election

At the federal election last year, Labor got a low return for its vote buck: 83 seats from 52.7 percent of the vote. By comparison, in 2004 the Coalition got 87 seats from the same vote, and in 2001 won 82 seats from 51 percent support.

Rudd will probably increase his majority at the next election, despite perhaps a small swing against. Incumbency tends to produce a seat-vote advantage, with a whole bunch of new MPs building a personal vote (what Malcolm Mackerras calls the 'sophomore surge'). In 1998, for example, the Howard government survived comfortably despite suffering a swing that the pendulum predicted would cost them office.

Most of Labor's 2007 poor vote-seat performance was in NSW - compare 2007 with 1993. 

Overlayed on this is the likely problem NSW Labor will be for Rudd in 2010. 

See the NSW extract of this table of cumulative swings 1993 - 2007. See how the outer suburban mortgage belt, eg Macarthur, is still very pro-Liberal relative to the last time Labor was in power.

And below is a mini-pendulum of NSW federal seats.

Labor's seven most marginal NSW seats were all Liberal-held before last November, so you can add a few points in 'sophomore surge' to their margins. (Probably not for Belinda Neal in Robertson; make hers negative.) On the other side, Macarthur and perhaps even Hughes must be vulnerable; Maybe, at a stretch, even Greenway.

Altogether, it suggests that perhaps Labor will cop a big swing, say 3 or 4 percent, in NSW but minimise the seat loss.

(People carry on about western Sydney being 'crucial'. I've asterisked seats that might lie in that area; as you can see the vast majority aren't in play.) 

All highly premature speculation, of course. I think also the state is due for another redistribution before then.

November 13 2008 New book on former PM

A book by Norman Abjorensen was launched last night. Here.

[irony on] A political force of nature The phone call came late one evening, from an old Liberal hand. 'Mate', he drawled, 'you c--ts have it all wrong. Rudd wasn't the reason we lost government last year. It was f--ken Swannie!' I heard ice tinkling. Then he hung up.

Astute political observers have long known that Wayne Swan is the Labor government's most lethal asset. One source sees in the Treasurer a forceful personality combining with the allure of danger: 'there's never a dull moment with Wayne.'

Another Labor warhorse puts it like this. 'The thing about Swannie is that when he mounts the podium he throws away his notes. He owns the space; it's a high-wire act. There's no-one else like him.'

Among Liberal ranks as well, the admiration runs deep. 'I wish we'd tried harder to recruit Wayne at uni', confides one frontbencher, 'he can turn a show around.' Another senior Liberal describes him as 'the most fascinating person in politics - for a very long time.'

And the prospect of a Swan Prime Ministership puts the fear of God into opposition members. 'Mate, if Swannie becomes [Labor] leader, it's all over Red Rover', despairs one. 'Then we're 'f---ed - for ten years or more. I'm f--ken leaving parliament!' 

Then he hung up as well.

[irony off] Tune here.

On remembering the Howard years

Caught a radio ad for the upcoming ABC series on Howard years, which had the former PM intoning something like this (from memory): 'When I came to office I had a clear idea of what I wanted to do and where I wanted to take the country'.

Ha ha. Anyone who was paying attention in 1996-7 will instead recall a new PM who, having promised before the election, in so many ways, not to change anything, now appeared not to know why he was in the job. (Apart from to settle a few old scores.) 

(Description also apt to the current PM, of course. Apart from the 'settling scores' bit.)

The 'narrative' evolved later, but is applied retrospectively. That'll happen with this lot too.

November 11 Newspoll says 55 to 45

In the Oz; tables here. Turnbull's approval up a smidgin to 53, his preferred PM down a bit to 22.

Trivia corner: Opposition leader John Howard (1995-6) never had approval (in Newspoll) as high as this one of Malcolm's; Kevin Rudd's (2006-7) was never as low.

November 10 My bets: next stop London

Regarding a bet made back when Kim Beazley was Labor leader.

And two out of three ain't bad.

[Note that under preferential voting Labour would almost certainly have retained Auckland Central.]

Joe in 2010? As you know, I don't believe Malcolm Turnbull will make it to the next election - unless it's an early one. Malcolm is possibly the most talented opposition leader in history, but my expectations come from a belief in the predominance of institutional forces over individual ones.

Julie Bishop's staff have been caught plagiarising once too often. Who else is there? Peter Costello might contend if the prize looks worth taking. 

And the big bear of a man Joe Hockey may be worth watching.

November 7 New Zealand election tomorrow

[Update: Mr Morgan has a further narrowing.]

Norm Kelly on the voting system in Inside Story.

Polls story here.

'Obama factor' If Helen Clark's government does substantially better than the polls (all still taken, I think, before the US Presidential result) indicate, perhaps we can put it down to an 'Obama factor'.

About 80% of people on this planet are somewhere between being rather pleased and ecstatic with the US result, and a bit of glamour might rub off on left of centre parties. For a few days at least.

If I were Ms Clark I'd mention the words 'Barack' and 'Obama' as many times as possible today. As in: 'I'll be meeting President Obama next year'. 'You never know, he might come to visit'.

US Presidential election wash-up: thinking points

• Half way down this piece the other day I noted that US polls are published with undecideds still included, which understates both candidates' percentage support. Given the absence of serious third party candidates, it was obvious Obama would get over 50 percent if he won by more than a few percent; it could not mathematically be otherwise.

It is good to publish the number of undecideds. But why not extract them from the results to give realistic percentage numbers?

• More white people voted for McCain than Obama. This white favouring of Republican over Democrat has, it seems, been the case at every presidential election since 1964. And it was only back then, during the Kennedy-Johnson years, that that race-party realignment took place. 

If you extract the 28 percent of Blacks and Latinos, you get a very conservative-voting electorate indeed.

Must learn more.

November 6 Me in Crikey on US election

Three this week: that Obama may not win (ahem) on Tuesday; dodgy stats early yesterday and a few facts - plus 'the Palin effect'- today.

November 5 The goose, as promised  But a happy goose. [Since removed from entry point, but is still available here.] Cheers.  Background to the goose here.

November 4 Please tell us, Gerard There is, of course, one person on the planet who already knows the outcome of the presidential contest, and that's Australia's own Gerard Henderson.

After every election, Gerard devotes a column to rubbishing people who got the result wrong, explaining that he himself always knew what would happen but did not want to ruin it for everyone else.

Ok, that's not quite fair; he sort of intimates that he rather suspected such an outcome. And I think last year's federal election was the first since the early 1990s which was not followed by such a Henderson column. (There had earlier been this.)

Anyway, here's me in Crikey yesterday, having a bob each way.

November 3 Picture worth a thousand words

Lesson 1: How to kill an opinion poll recovery, increase your opponent's support and flatten your own. (Graph from pollster.com)

October 30 pm New York's November 4 ballot paper

Here, from Josep Colomer's blog. 

As you can see, lots of positions voted on at once. Good. But very user-unfriendly and confusing. McCain-Palin listed three times and Obama-Biden twice. !!

Just looking at it hurts the head; can understand why so few people vote over there.

October 30 Newspoll issues: economy and national security

In the Oz: yesterday's issues tables (here's Mr Shanahan's take). The right hand tables' labelling is confusing; these ones from the Newspoll site clear it up.

To me, the most interesting things are the economy and national security. They're heavy issues that incumbents will generally be favoured on. (They were the outer layers on Mr Hartcher's beloved Babushka doll.) It is a cart and horse thing: they're favoured largely because they're in government, not vice versa.

On top of that you can kind of overlay a general advantage for the Coalition in perceptions of economic management - but not (one would think) national security.

Since the change of government, perceptions in both have moved big time towards Labor. But the government is still not actually favoured; the Coalition's leads have shrunk from 24 and 23 in October 2007 to four and five this month. Perhaps it takes a while for these things to wash through.

While the Coalition is favoured over Labor on the economy, Kevin Rudd is greatly favoured over Malcolm Turnbull on economics, and - one would guess - Wayne Swan over Julie Bishop. The Coalition's advantage on the economy will perhaps/ probably disappear next time it's measured. 

But Labor's deficit on national security has actually increased since February. Is this because the issue (or what the phrase means to respondents) has dipped below the radar?

If so, it might take some activity in national security for the government to take the incumbent's advantage.

October 29 The Australian versus Kevin Rudd: treat 'em mean, keep em keen?

Without entering into the rights and wrongs of the issues(!), everyone would agree the Oz is tougher on the Rudd government than any other major newspaper. Yet that paper seems to get most good government scoops. 

Mr Rudd needs a decent script editor

The scoops are good for the government too of course. This one last week depicted Kevin Rudd in important conversation with the world's most powerful person. These sorts of stories were routine under John Howard of course, but this account had our hero having to explain to President Bush what the G-20 is. 

That sounded a little too good to be true, and has been pooh-poohed by Bush people (half-way down this, 'G-20 Maneuverings'). Rudd has back-tracked.

There were also elements of unbelievably in a recent George Megalogenis piece about the PM in talks with public servants on how to handle the financial crisis. Stuff like this:

"Rudd kept coming back to the same point: What is the most responsible thing to do?" say senior sources who are familiar with the discussion. "Forget about what is going to win us the news (cycle) or lose us the news (cycle), what is the most responsible thing to do?"

Have we really reached the point where the PM needs to say such things to public servants? Or have the youngsters in Rudd's office - one of whom was presumably George's source - watched too much 'West Wing'? 

It is often said that Mr Rudd needs a good speech writer. Let's get him a decent script writer/editor too.

And perhaps a larger question is: what are journalists supposed to do when presented with such clunky storylines and dialogue?

October 28 Newspoll says 54 to 46

In The Australian, tables here, Dennis here. Sky-high approval for the PM, and high approval for the opposition leader. 

On the ETS, Newspoll's questions are different to previous ones (on 1 July and  9 September) and so it is not really possible to say whether attitudes have changed.

October 27 The Palin fairytale continues ...

Let's codify a political rule of thumb: candidates who send a party's true believing, one-eyed base into cartwheels of exultation should be avoided. They tend to embody partisan fantasies about what the electorate wants. 

Recall our own Mr Sheridan's rapture early last month. Recall also the Latham experiment. 

And the idea that Governor Palin can be seriously in the running in 2012 is crazy. Ronald Reagan - the non-establishment candidate who beat the establishment one (GHW Bush) and then the incumbent in 1980 - she ain't.

Except in the fevered minds of true believers.

October 24 NZ election in two weeks

Morgan's latest has Nationals way ahead of Labour (43 to 32), though not enough for majority in own right. Greens on a high 11.5.

Pollbludger has collected recent opinion polls, and the standout feature of most of them is that the Nationals look set to romp home. Except ... which other party/parties would support a National government? Perhaps an ACT member or two?

October 22 US election: calling all maps

As a mere dabbler in US electoral politics, I have only recently looked at polling/state maps such as pollster.com and The Oz . Electoral-vote has the advantage of having each State's last several election results.

If people send me some more good map links, I'll put them up. [Update: more at uselectionatlas, fivethirtyeight, our own ABC and CNN.] [And Newshour.]

October 21 US presidential election: hold your horses

Having prepared psychologically for what I reckon is a slightly more likely than not Obama victory, I must again protest against the widespread expectation that he is a shoo-in.

The current $1.08 to $7.50 payouts at Centrebet are crazy, in my humble opinion.

Journalists almost invariably refer to Obama's 'double digit leads', but that's not true. This great site (via pollbludger) has at time of writing a 'trend' lead of about six. [Here's The Australian's map.]

Opinion poll reinforcement

I suspect the unprecedented amount of opinion polling is influencing expectations. There are now so many of them, in so many States, saying the same thing, day after day, that the expectations of a Democrat win are multiplying. But they are all (on average, anyway) saying that Obama is only a few points ahead.

Not quite a pattern

Here are a couple of instances from the past - too few to constitute a 'pattern'. I don't know whether this has been noted before. In 1988 and 2000, after the last two two-term presidencies, the candidate from the incumbent party trailed very badly before making lots of ground in the final weeks/months. In the case of GHW Bush it was enough to win easily, while Al Gore only managed an approximate dead heat (vote-wise) on polling day.

Of course, a difference is that Ronald Reagan and Bill Clinton were still very popular at the ends of their presidencies, while GW Bush ain't. 

But still, if McCain deep-sixes his veep candidate for two weeks and distances himself even more from Bush, then if he's lucky something might take people's minds off the economic outlook, maybe - you never know your luck - on the final weekend Osama bin Laden will endorse Barack Obama, as he did John Kerry in 2004.

And if there is such a thing as the Bradley effect, then six points is nothing.

Forget the Paddy Power stunt. Fat lady is still squeezing into her corset.

New Yorker article on the Australian ballot (Re my other hat.) By Harvard historian Jill Lepore, this is a great piece.

October 20 Federal voting intentions

Nielsen says 56 to 44. Tables here.

Federal repercussions of NSW by-elections?

Only one: there is no adjective strong enough to describe how NSW voters will feel about its government in 2010.

October 19 ACT: Scenes from an election

Here are some snaps from Lyneham booth at yesterday's ACT election. Click any for larger version. Note the palm pilots - as opposed to big books - used to identify voters and tick them off, in the third photo (bloke in green t-shirt). [Update: Canberra Times.] 

October 18 Malcolm Mackerras's predictions

I emailed Malcolm asking for his ACT election prediction, 'on or off the record'. This was his reply:

'On the record for the ACT I say seven Labor, six Liberals and four for the Greens.  

While I am at it I may just as well make on the record predictions for the NSW by-elections.

I say the Liberals will gain Ryde, the Nationals will gain Port Macquarie and Labor will retain Cabramatta and Lakemba.

Cheers,

Malcolm'

I think I saw the Canberra Times's Crispin Hull on telly last night predicting 7, 7 and 3. 

The paper, which isn't updated online until around 9am, apparently has a Patterson poll today showing a 'late swing to Labor' from a fortnight ago. First instalment of same survey yesterday had the opposition leader going backwards on personal ratings.

[Update: they don't update on Saturday's, it seems. Here is a photo of the front page, which includes the below prediction from the poll:]

• New blog the Tally Room predicts: 'ALP 7, Liberal 6, Greens 3, Richard Mulcahy [a former Liberal] 1'.

October 17 My bets: money for jam in Ryde?

A Liberal win in NSW seat of Ryde currently pays $1.60 at Centrebet. I found those odds irresistible. (But I never recommend betting.) [3:25pm: down to $1.20 (v $4.00 for Labor).]

In other news, John McCain now pays $7.00(!) and the Libs in ACT $4.35.

Speaking of McCain ...

Cometh the hour, cometh the goose I now believe it more likely than not that Barack Obama will be elected US president on November 4.  If this happens, it will then be incumbent on me to post a big goose here.

The goose will signify me, for my regular, unequivocal predictions of the opposite. 

I shall largely cop it on the chin, although being human will present some mitigating factors - the Palin fiasco of course (I do believe her appointment is the main reason Obama is currently so far ahead in the polls) as well as the Wall Street meltdown. 

I have probably also under-estimated the degree to which Americans really mean it when they talk about 'change'; Australians are more timid. And am surprised at how unimpressive McCain has been; had been thinking of the 2000 version.

But it is true my predictions included no provisos.

Still, it's not over yet. There are three weeks to go, and anything can happen; it will surely get uglier. But am prepping the goose for the somewhat better than even chance of an appearance.