Tag Archive for 'psephology'

The truth of polls and the epistemology of politics

Published
by
Mark
2 weeks ago
in Elections, Media, Polls, Queensland, Sociology, State/Territory Elections and Victoria
. 13 Comments

The Poll Bludger has the numbers on the latest Nielsen poll for Victoria. Labor leads on the 2PP 55-45.

The Age trumpets this result as Victorian Labor “defying the national trend”. No doubt other papers are saying the same - I haven’t looked.

I’ve been arguing for a while that there is insufficient evidence to conclude that there is a national trend against Labor, and that in fact thinking about disparate polls in seven different jurisdictions with differing political histories, cultures and current circumstances as constituting a trend makes little sense. My contention for a long time has been that elections are unrepeatable and singular events and that epistemologically we can know much less about electoral behaviour and find grounds for prediction with much less certainty than we think. Political behaviour follows few laws and a lot of conclusions reached after the fact are questionable.

But there is a sort of reflexivity feedback loop built into the way we think about politics and the way polls are reported. Particularly at state level - where polls are few and far between - one poll which struggles to form a series can have a large impact on perceptions, and thus the interpretations of the public and the press and the morale of politicians and “momentum”.

Continue reading ‘The truth of polls and the epistemology of politics’

Truthiness versus Truth II: Now with graphs!

Published
by
Mark
2 weeks, 3 days ago
in Foreign Elections, Sociology and USA
. 6 Comments

More on the instant revisionism from the Republican Noise Machine in the wake of Barack Obama and the Democrats’ victory - this time scatterplot and red state blue state rich state poor state make a graphic point about the claims that the Republicans’ loss was somehow artefactual. It’s worth adding that the problem of the under-representation of Democratic votes in terms of seats adduced also goes to the horrendous architecture of the American political system - entrenched and partisan gerrymandering in many states, the two party monopoly, disenfranchisement and appallingly conducted elections, and all the other factors which distort popular will and poorly represent it.

Continue reading ‘Truthiness versus Truth II: Now with graphs!’

He’s tanned, rested, and ready to lead!

Published
by
Kim
3 weeks, 3 days ago
in Media and Polls
. 27 Comments

Newspoll has Labor on 55% 2PP (up 1 point, within the MOE). Kevin Rudd is on 62% (up 3) on the PPM to Malcolm Turnbull’s 22% (down 3).

Brendan Nelson peaked at 16%.

The Opposition Organ says:

But it remains substantially higher than his predecessor Brendan Nelson.

How substantial is substantial, I wonder?

Continue reading ‘He’s tanned, rested, and ready to lead!’

Labor states on the nose!!!

Published
by
Mark
4 weeks ago
in Federal Elections, Media, Polls, Sociology and State/Territory Elections
. 58 Comments

I’ve written before about why I think that the “media narrative” masquerading as psephological analysis that there’s some sort of automatic fall in support for state Labor parties because of some putative desire among voters to have different parties governing at different levels. I suspect the proximate origin of this meme is actually the “wall to wall Labor” scare the Liberals ran in last year’s election. As Kim was suggesting the other day, this is a political tactic that normally indicates despair among incumbents, and it’s completely wrong to assume it reflects some sort of psephological law or reality in voting behaviour. Anyone familiar with the political science literature knows that attempts to demonstrate any posited strong correlation between state and federal partisan choices over time falls down very quickly - even in New South Wales where it’s long been political myth that the strongest case can be made. In part that’s because there are two few cases of actual partisan change in elections over much of the postwar period - something that becomes immediately apparent when you think about the fact that the federal government didn’t turn over at all between 1949 and 1972.

There was another outbreak of this guff on Monday in The Australian, apropos of quarterly newspolls in New South Wales and Victoria. There was the usual news story and two op/ed pieces to ram home the point. It was intriguing to see the frame applied to the comments of ABC election analyst Antony Green: Continue reading ‘Labor states on the nose!!!’

US election: the demographics

Published
by
Mark
4 weeks, 1 day ago
in Culture, Foreign Elections, Polls, Race, Religion, Sociology and USA
. 22 Comments

The big note of caution should be that this data is drawn from exit polls which don’t take into account the very large early vote, and that exit polls have certain problems of validity and reliability.

However, there’s some interesting reading at both RedBlueRichPoor and FiveThirtyEight. I suspect the absence of the early vote in the sample diminishes the increases in youth and African-American turnout. And it’s just as well to remember that the election was actually reasonably close in the popular vote, and that the electoral vote map has been tweaked rather than transformed. Trends which were operating this year need reinforcing through the upcoming electoral cycles, and much will depend on Obama’s performance in government. You can never claim that there’s a realigning election on the basis of one cycle (06-08), though my bet would be that this will turn out in retrospect to have been one, for a whole host of more qualitative and interpretive factors.

I haven’t seen any comprehensive data as yet on Hispanic and youth votes’ partisan composition, but some evidence presented during the CNN coverage suggested very strong 60+% Obama votes in both populations. It’s worthwhile noting that the demographics of many states which are swinging to the Democrats strongly - or showing signs of such swings (ie in Texas, Arizona and South Carolina) - are fast changing to a much more heterogenous composition. And that generally, the younger cohort is more socially liberal and less hung up on racial barriers and all the baggage of the culture wars - including among younger Evangelicals.

There’s a lot here for Obama and the Democrats to make hay with in terms of partisan and geographic realignment, if they play their cards well.

Continue reading ‘US election: the demographics’

US election: Yes we can!

Published
by
Mark
4 weeks, 1 day ago
in Activism, Blogging, Elections, Foreign Elections, Media, Polls, Sociology, Technology and USA
. 16 Comments

Image of spontaneous street celebrations in Harlem courtesy of matt semel at flickr - reproduced under a Creative Commons licence.

No doubt one of the big stories about the US election will be the influence of the blogosphere and the netroots. In many ways, the rise of the intertubes in politics was an unintended consequence of the Rove approach to politics, as Publius perceives:

The bigger story is that this same anger – this same frustration – has led liberals to organize in more numerous and consequential ways. In the last few years, we’ve seen new think tanks. We’ve seen blogs flower. We’ve seen the rise of media sites like TPM and Huffington with real journalistic chops. We’ve seen unprecedented efforts to register and canvass voters.

In short, we’ve seen a new energy driving liberals back to politics.

In an opinion piece at ABC Online, Barry Saunders sums up the changes that net based activism and citizen journalism have wrought:

The impact of social media on this election has been enormous. Whoever takes office will have to deal with widely available factchecking data, embarrassing videos, rabid wingnuts, opinionated bloggers and TV hosts, and a massive number of new voters and donors who feel they have invested in the American political process - as well as two wars and a collapsing economy. Here’s hoping they know what they’re doing.

Continue reading ‘US election: Yes we can!’

The American Possum (and the Australian Possum)

Published
by
Kim
1 month, 2 weeks ago
in Blogging, Elections, Federal Elections, Foreign Elections, Politics, Polls, Sociology, Technology, The Web and USA
. 5 Comments

When I first discovered FiveThirtyEight.Com (courtesy of Down and Out of Sài Gòn on one of the American election threads here at LP, if I recall correctly), I thought of Nate Silver as America’s Possum. Coincidentally, there’s a profile of Silver published in New York magazine which makes some very similar points about the emergence of a statistician doing psephological wonkery as an avocation into a major source of expertise and information on elections as an article sounding that theme from Monash University Journalism Professor Chris Nash in the new edition of the Pacific Journalism Review.

Just think what elections would be like without Possums and Silvers! If all we had to rely on for psephological goodness was the dead tree media…

Government moving too slowly on IR; Essential Research 57-43

Published
by
Mark
2 months, 1 week ago
in Industrial Relations and Polls
. 16 Comments

…45% of Australians think so, according to this fortnight’s Essential Research poll. As a bit of an addendum to my earlier post about Julia Gillard’s speech last week to the National Press Club on the detail of the Forward with Fairness bills which will shortly be introduced into parliament, I should also note that many Labor MPs have been concerned by reports they’re receiving from constituents about continuing abuses of workplace power. This is more the everyday bastardry that WorkChoices encouraged, rather than the headline anti-union moves of big corporations like Telstra. A lot of voters assumed that WorkChoices had already been “torn up”, and there’s significant pressure on Gillard to bring forward some of the implementation dates for aspects of the new legislation.

The whole “keep business satisfied” implementation agenda might have seemed like a good idea last year. It’s not looking so flash now, particularly as the ACTU finally wakes up to the fact that they’ve effectively been locked out of the policy making process.

Elsewhere: More discussion of the poll at The Poll Bludger. Also interesting is the comparison with ratings of attributes between Malcolm Turnbull and Kevin Rudd (with the proviso that the data on Rudd dates from June). Turnbull will be worried at the 47% “out of touch” figure. How do you actually turn that around? Brendan Nelson didn’t do so by emoting and going trucking.

B… b… bounce?

Published
by
Kim
2 months, 2 weeks ago
in Media and Polls
. 18 Comments

Oks, there’s got to be some way to work this into a drinking game. Just to prove that political tragics are rooly cool like the kidz on West Wing and not strange nerds really.

Possum has enabled a feature on his blog where you can guess the size of any Newspoll bounce that might occur now that Malcolm Turnbull is Oppo Leader. Wisdom of crowds and all that. No mention of any prizes (hint! hint!*) - at least you can make a buck from the betting markets…

So saying, the average Coalition bounce from a leadership change was 5.7% on the primary and 6% on the TPP. This compares to Labor’s average leadership change bounce of 3.6% on the primary and 1.7% on the TPP. The total average leadership change bounce was 4.6% on the primary and 3.8% on the TPP across the 6 historical examples we have available.

And I’ve got my own idea for a bit of crowd sourced political commentary. Let’s pretend we’re the press gallery and set a % which, if not met, will be decried as a setback, or completely unrelated and probably because of the current position of the moon in the lunar cycle, depending on what mood Dennis Shanahan et al are in. And then we can set a % sufficient to ensure the production of headlines like “Australian politics has fundamentally changed”, “the honeymoon is now over”, “Turnbull reinvigorates Coalition” etc, etc.

Be your own press gallery. Write your own political narrative.

Ps: If Turnbull can’t get 5% on the 2PP, he’s toast. Peter Costello is willing to be drafted. ;)

*Update: Now there’s a prize!

Climate change and electoral politics

Published
by
Mark
2 months, 3 weeks ago
in Climate change, Federal Elections, Industrial Relations, Polls and Sociology
. 18 Comments

There’s lots more interesting stuff in this report at Australian Policy Online about two exit polls taken at the time of the 2007 federal election (and the AES), but this might be a relevant thing for Kevin Rudd, Penny Wong and the ALP to remember in the context of the emissions trading scheme and international negotiations on climate change response:

Industrial relations and global warming were key issues for the Labor voters who took part in all three polls, with two of the polls revealing that global warming was the prime concern among voters who changed their vote between the 2004 and 2007 elections.

Newspoll: Labor 56-44

Published
by
Kim
2 months, 3 weeks ago
in Polls
. 8 Comments

Via The Poll Bludger, the same old same old from Newspoll. I think what Rod Cameron had to say on Lateline had a lot of merit - voters will give Kevin Rudd at least another year before making a final judgement. In the meantime, the Liberals probably are wasting their and all of our time with their constant shenanigans. Their best bet would be to lie low and stop making themselves the story. They’re unlikely to, of course. Too much a bunch of prima donnas, plotters and media tarts. And now we’ve got this to look forward to:

But you can be sure that from the end of the next week until at least the end of the following week, it will be all-Costello-all-the-time in the Australian media.

Lord save us.

This question from Essential Research (which incidentally has Labor sitting on a 59/41 2PP split) is a hypothetical, but it’s still an interesting result:

Kevin Rudd is preferred to Peter Costello head-to-head 53 per cent to 27 per cent.

Elsewhere: Possum shows why the Costello dithering is doing a lot of damage to the Libs.

The middle-classing of Labor?

Published
by
Kim
3 months, 3 weeks ago
in Federal Elections, Howardia and Sociology
. 40 Comments

The graph is from page 101 of a Parliamentary Library paper on the 2007 federal election. It shows voting patterns disaggregated by electorates ranked in four socio-economic groups according to income.

As Brian Costar observes at Australian Policy Online, it doesn’t mean what you think it means - particularly if your perception of “Howard’s battlers” is that they’re all outer suburban. It’s an artefact of a correlation between voting patterns and low incomes in rural and regional electorates.

I haven’t had a chance to look at the full report, but I’m sure there’s lots of interesting stuff in there for the psephologically inclined.

Reality and unreality in the pundits’ world

Published
by
Kim
4 months, 1 week ago
in Books, Writers & Writing, Climate change, Economics, Energy, Environment, Immigration, Media and Polls
. 17 Comments

Let’s take a look at today’s political “news”, News Limited style, and the ongoing construction of the “media narrative” that according to the press gallery gang, is the only news fit to print.

As noted here, The Opposition Organ spent a bucket of dosh to add extra questions to Newspoll, and chose to run with “Voters Want Costello” as its front page headline over the (presumably less welcome to the masthead of denialism) numbers on climate change, showing overwhelming majorities attributing climate change to AGW and support for an ETS, with a big majority for “not waiting on the world”. So that’s establishing the news agenda through polling to feed the current “media narrative” - centring on the Liberal leadership and Peter Costello lovin’ in particular. And selectivity in emphasis. Then we get selectivity in reporting. The numbers in Newspoll, as Possum points out, don’t show that the voters the Liberals need to persuade are particularly persuadable by a putative Costello return:

The Coalition needs ALP voters to shift to the Coalition, yet ALP voters have a breakdown of 15% more likely and 20% less likely. If Costello became leader, he might not lose voteshare, but neither does he look like he would gain much based on these results.

But Dennis Shanahan doesn’t mention that.

Let’s go back a bit and remember, as Mark pointed out in his review, that the extracts from Inside Kevin07 that kicked the Costello talk off were themselves highly selective - one bit of research done before Rudd became leader and highlighted while the other internal polling and focus group research showing Costello for PM being about as appealling as a piece of wet lettuce was studiously ignored. And let’s not forget either that the “Costello the Saviour” narrative basically depends on the publication date of a book! Leadership calculation by publishing schedule! Melbourne University Press and book distributors hold the nation’s future in their hand!

Then, the big showdown Bolta talked up on the Coalition’s emissions trading scheme stance comes - and Nelson gets rolled.

Meanwhile, the Labor government has basically done away with mandatory detention.

I would venture to suggest that is rather more important than all this other confected nonsense.

Continue reading ‘Reality and unreality in the pundits’ world’

The state of political blogging II

Published
by
Mark
4 months, 1 week ago
in Culture, Federal Elections, Media, Politics, Polls, Sociology and The Web
. 2 Comments

Last year I shared some thoughts on the state of political blogging in Australia. Trevor Cook has just examined the claim that the blogging phenomenon is “losing impetus”. I’m not sure that’s so, and coincidentally, I’ve just sent off a write up of the talk I gave at the Public Right to Know Conference at UTS last year, for a special issue of the Pacific Journalism Review being co-ordinated by the Australian Centre for Independent Journalism. You can read it here [link to pdf].

Continue reading ‘The state of political blogging II’