This is the name of a new academic paper (still in working form) that Aaron Martin and Dr. Juliet Pietsch are currently writing – and they were extraordinarily kind enough to let us have a little bit of a peek before it gets published. The abstract of the paper lays out the broad contents and purpose:
This paper examines to what extent older and younger generations of Australians contribute to or disrupt the stability of the Australian party system. The Australian and international literature would lead us to expect that older generations contribute to the stability of the party system in a number of ways: by being less likely to vote for minor parties, having a stronger partisan identification and being socialised into a voting pattern at an early age. Using data from the Australian Electoral Study this paper will test these expectations by looking at generational effects. In doing so we find that older Australians contribute to the stability of the Australian party system by being less likely to vote for a minor party and being more likely to vote for the same party over time. We also find that young people have a weaker partisan identification and enter the electorate less likely to vote along parental lines. Controlling for a variety of factors we find that generational factors are strong predictors of voter volatility.
While that is extremely interesting in and of itself (and it really is, the research is a cracker!), where the paper really pounds into the pointy end of modern electoral politics is through their analysis of intergenerational voting behaviour.
One of the themes we regularly explore here on the blog is the notion of the Coalition’s Demographic Train Wreck – how historically the Coalition have received a premium level of voteshare from the generation born before World War 2 (that we call Gen Blue), how that vote has delivered them electoral benefits for over 30 years – but it hasn’t been replaced in the following generations, leading to a structural decline in the Coalitions vote as this older generation continues to become a smaller proportion of the total electorate. Our most recent piece on this can be seen over here.
Our work on this topic was largely derived from the excellent research undertaken by Ian Watson (whose historical age-cohort Newspoll data we use regularly). The most recent update of his paper, “Is demography moving against the Coalition? Age and the conservative vote in Australia, 1987 to 2007” can be found on his site.
The Grey Vote: Ageing and Cohort Succession adds a new dimension to this argument by not only using the Australian Election Study to track the primary vote of 4 generations of Australian voters (Depression Era/Boomers/X and Y) since 1967, but also delves into issues of partisan identification over time, increasing minor party vote levels, the declining influence of parental voting patterns on a persons vote and how this all washes into the potential for future electoral volatility.
With permission (aren’t they just fabulous), we can have a look at how some of this is playing out. The first thing to note here is that the data comes from the Australian Election Study – which is a large post-election poll undertaken in the weeks and months after every Federal election. The one problem that the AES has though, is that it tends to slightly inflate the vote for the party that won the election, and slightly deflates the results of the major party that lost – but only by a few points at most. In some respects, this is almost unavoidable as it’s an extremely common occurrence for these types of “post event” polls and goes to the psychology of a small proportion of the public either liking to be on the winning side, or not recalling their actions and simply going with the majority.
So saying, the patterns involved here are consistent across both time and differing political flavours of government administrations – so the trends are robust.
First up, using the data in the paper, we’ll track the four generations of voting behaviour for the Coalition primary vote. (just click to expand the charts)
As we can see, those born in the Depression Era have a consistent and considerable pro-Coalition voting disposition compared to any other generation. Also worth mentioning is how each subsequent generation consistently votes for the Coalition at smaller rates – with Depression Era being strongest for the Coalition, followed by the Boomers, followed by Gen X and Gen Y coming in weakest.
You might notice that the Gen X result in 2004 appears to be unusual? That’s because it is – I have it on good authority from multiple sources that the Gen X primary vote for the Coalition in 2004 is about 7 points overcooked in the Australian Election Study compared to what actually happened. However, don’t let that detract from the rest of the data – having one outlier of this magnitude isn’t statistically unexpected in a set of 32 observations in a post-event style poll.
If we now move on to the ALP primary vote:
What we see is that the only generation that votes in a consistently different manner to the others here is the same Depression Era cohort – voting in smaller proportions for the ALP than any other cohort at every election (including the 2004 election once we take into account the outlier nature of the Gen X 2004 result). Boomers, Gen X and Gen Y however all seem to broadly move together, and at the same rate for the ALP over the last 20 years.
What we have started to see, and expect to continue to see at an increasing rate in the future, is the national primary vote of the Coalition losing the premium that the Depression Era voters have historically delivered them – as the Depression Era generation becomes a smaller proportion of the electorate – resulting in a national primary vote that increasingly approaches the average value of the other 3 generational cohorts.
How important is that Depression Era premium vote worth for the Coalition?….. I hear you ask.
Well, if the Depression Era vote was removed from recent election results, Howard would have lost in 1998, lost in 2001 and have been line ball in 2004 – it’s the most electorally significant demographic for the Coalition and one whose electoral power isn’t being replaced in other generations naturally – meaning the Coalition will have to attempt to replace the natural vote boost they receive from their Depression Era voters with other generations, deliberately via some political strategy, should they wish to be regularly in government again in the future.
The way the demographics are changing doesn’t mean that the Coalition will never be in government again – but rather, that without solid generational repositioning they will far more times than not, find themselves in opposition. Ultimately, however, their electoral future is in their own hands.
The non-replacement of the Coalition vote premium links us in to another fascinating phenomenon that Pietsch and Martin unearthed in their research – the way that the partisan identification of ones parents is having decreasing influence across time on any given person’s vote.
The Australian Election Study asks:
“Did your father have any particular preference for one of the political parties when you were young, say about 14 years old? And what about your mother?”
When the responses are combined and compared against how the respondent actually voted in any given election, and then broken down by age cohort, we get a clear trend among all 4 cohorts:
Only a tad over 30% of the population now vote according to the partisan identification of their parents and it’s been flat for 3 elections now. It’s as if we’ve entered a relatively new post-parental partisan voting era that is being manifest in all generational cohorts.
In the literature, one of the best (historically at least) predictors of the way a person will vote is the way their parents have voted in the past – yet here we see this relationship effectively breaking down in Australia. Also worth mentioning here is another piece of related data unearthed in the paper – the proportion of respondents that have always voted for the same party (although this time, we’ll leave Gen Y out since they havent been voting for long).
Rather than this breakdown being caused by children all voting for the major party opposite to their parents, or people always voting for a major party – a significant proportion have been giving their vote to minor parties. Pietsch and Martin also tracked the generational minor party voting patterns across time.
What stands out is not only how the Depression Era generation have the lowest rate of minor party voting, but how – generally – the younger a person is the more likely they are to vote for a minor party.
With a structurally declining Coalition vote, a quite variable ALP vote and an increasing minor party vote, there is a growing potential for volatility and decreasing political stability in two party government terms as we head into the future if these trends maintain themselves. That is a topic that the paper explores in some depth and is well worth reading.
The other serious issue the paper explores is the robustness of these generational effects through some quite clever multivariate regression models that control for other variables such as income and education. They find statistically significant generational effects with the Coalition and Minor Party vote.
As it’s still a working paper, it’s not yet ready for publishing – so a big thanks goes out to Aaron and Juliet for allowing us to have a sneak peek at what the paper will contain. When it eventually gets published sometime over the next 9-12 months or so, we’ve been promised a copy so you can all have a read – it will be worth the wait… it’s an absolute cracker.
35 Comments
Fascinating.
A possible flow-on effect that springs to mind would be a reduced meaningfulness of seat margins. If voting patterns and partisan identification become increasingly volatile, surely your margin would become less stable because of the increase in swing voters.
Fantastic research.
The “vote the same as the parents” is especially interesting.
Very interesting and very kind of them too!
Even with the gen blue – Howard was dead lucky to win in 98, given the cliff down which the vote fell in all cohorts.
If only…..
I agree with grog that the ‘vote like dad did’ is fascinating. I had long ago assumed it was set in stone that everyone voted liek their parents. Except for the bit that I don’t.
Us oldies are virtually set in our ways and it would have to be something resembling a natural disaster (like for instance the Dismissal,or WW2) for me to change from my mainstream party. However my siblings are not so set and a general dismay at the majors is likely to push one of them to the minors, the other one has however remained true.
Good stuff.
Would extrapolating to 2010 be meaningful?
Would love to see a time series of generational voting power, with forecasts. Ie, when can us Gen Xers expect to wield the gauntlet?
Can you oblige, Possum?
kymbos – you mean when bb replace gen blue completely?
given the info at hand – the coalition as it is better get used to being in opposition – they might face a Menzies era in reverse
Actually, I just looked again at my projections of what 1949 – 2007 elections would have been had we had a system where number of seats won were more closely associated with the actual vote percentages, and it turns out that the current system having previously been overwhelmingly to Lib/Nat advantage, is now actually favouring the ALP, and a more proportional system would now help the Libs more than the current system.
Will we now hear calls from the Libs to reform the electoral system to a more proportional system?
VP, no. This is stuff that shows up over decades. In any three year election cycle, the shift caused by a new 3 year cohort of 18-21 years olds joining the voting population while the oldest groups sees people die will always be small compared to the ebb and flow of short term politics and issues.
The 1967 and 1969 AES studies showed the transmission of party loyalties to children was still strong. The 1979 survey was the first to reveal this transmission breaking down, especially in middle class Liberal voting families. Each AES study since has shown transmission weakening. That is not surprising given social mobility is on the rise.
Most western countries have the same political parties they did around 1920. This is largely because once in existence, most parties have been able to survive in a changing world by altering their image and message. On the surface Australia’s party system still has the same class basis as in 1920, with three parties representing Labor, Urban capital and rural capital. Our party system carries on the shadows of a long gone class party system even though the society we live exhibits a level of social mobility that would have been unimaginable in 1920.
David, there’s a simple reason for your finding. From 1955 to 1975, preferential voting favoured the Coalition because the only significant minor party was the DLP which directed preferences to the Coalition. From 1975 to about 1987 preferences split evenly and since 1990 they’ve favoured the Labor Party. The advantage/disaadvantage you’re finding is caused by preferences and the nature of minor parties.
Antony – yes – the two party system has that effect, and it did serve the conservatives well, but now it may actually be in their interests to move for change.
agree re the outdated classisist basis of the 2.5 major parties and it being totally irrelevent to the modern era
another reason we should ditch the two (and a half) party model in favour of something more aligned to the 21st century
You said proportional though. They’re actually debating optional preferential voting. It’s not the proportionality of the system, it’s the importance preferences.
As the proportion of major party vote declines, optional preferential voting will be the first port of call before PR. Compulsory preferential voting can seriously distort the translation of seats into votes in multi-party systems.
either way – true proportional as in the senate, or OP, the result would be something where the number of seats won would more accurately reflect that party’s level of support
do you think the Libs would make a push for OP?
‘we’ should ditch? You don’t ‘ditch’ party systems, they evolve. Party systems only get ‘ditched’ by major social upheaval. Many European countries changed their party systems after the Second World War, but that sort of social dislocation is rare. The end of the Cold War caused major changes to the party systems in many western European countries, such as Italy and all of eastern europe. Some countries like Canada and Belgium have had their party systems completely re-structured, but in both cases that was caused by the assertion of linguistic differences over more traditional class and church-state party divides. European countries like the Netherlands and Denmark saw the rise of radical liberal parties in the 1960s, a similar trend eventually producing the Greens. Only Italy of Western European countries has taken the parties, scrambled them and re-arranged them.
Yes. Some already are. The NSW Liberal Party is strongly behind OPV. The Liberals and Nationals don’t run against each other anymore, so the entire reasoning about why it was adopted back in the first half of the 20th century has disappeared.
And boy, didn’t that work out just peachy for them!
Imagine trying to write “The concise political history of post war Italy” – it would be a multiple book set thousands and thousands of pages long!
It is actually easier that you think to paint the broad themes of 1945 to 1990. One third of the electorate voted Communist, but neither NATO or the Vatican would countenance any party forming a Coalition involving the Communists. So you had a shifting array of minor parties orbiting a dominant Christian Democrat party. When the corruption bred by permanent government combined with end of communism, the system fractured.
Possum,
I blame it on the Bossa Nova.
http://www.youtube.com/watch?v=TIGwKUqbTGc
Very interesting figures indeed, and thanks to Aaron Martin and Dr Juliet Pietsch for giving us this sneak preview. Ditto for Poss.
My grandmother was born prior to the great depression and she fitted the conservative voting pattern to perfection. A solid working-class supporter of any old candidate the Liberals chose to field. It just wasn’t done to mention the Labor Party to her without hearing “They’re nothing but a bunch of Commos, unionists and traitors to the country” was one of her less offensive descriptions. Aged about fifteen I asked her why she voted for the people who had done the least for the working-class. A torrent of Italian-via Corsica-language meant she wasn’t prepared to answer the question. Minor Parties were unspeakable, especially the DLP. Her daughter-my mother-followed the same pattern, and the day I admitted I voted Labor was rough.
Perhaps the present Libs wanted another savage recession a few months ago. A last ditch effort to get more voters.
Poss mentions education or by income as interesting ‘multivariate regression models’ which I’m not sure what that means. Could the choice of industry a person chooses to work in could be interesting. This could be at marked variation to earnings.
Just a thought. Thanks once again Poss.
venise – given that, as Antony made clear once again, the DLP kept the Libs in power for quite a long time, your granny’s animosity is hard to fathom
Also, that nice catholic anti-commie Bob Santamaria was influential in the founding of the DLP.
David: this was to do with the fact the DLP was a Catholic Party. Family legend had it that back in Corsica the local priest had attempted to interfere with her-as my mother put it. She was extraordinarily beautiful, even as a child. Great grandpa was less than amused; and a man who loved horses and who would never have taken one to a horse, took a horse whip to the priest and thrashed him. After that there was total contempt for Catholics. One didn’t dare mention the word to her, and all of us were atheists.
Mother married an Anglican who promptly fell into line.
I am painfully aware of the sinister Bartholomew Augustine (call me Bob) Michael Santamaria. He filled me with the same hackles-rising hatred that I feel for Phillip Ruddock.
yes – he was a nasty piece of work old BA
sadly, there are still gargoyles like him around in politics/media today
ah well…
seems like unless the Libs get smart PDQ, they will spend a long time in the wilderness
This is a wonderful thread When you read analysis and discussion like this you understand why the mainstream media is losing ground. News Ltd has access to these numbers and yet what do we get Shanahan’s delusional ramblings.
It’s a pity there are no state equivalents to the AES, as arguably demography combined with the different issue base made Labor the ‘natural’ party of government at that level by the 1980s.
I’m with you there Andrew – a state based AES would just be marvellous.
I think it’s arguable that national security at the federal level plays into helping the Libs enormously in terms of broader trust perceptions of a political leader – to the point where it saved Howards bacon in both 2001 and 2004. Between that and the dynamics of larger federal electorates compared to State ones, I reckon it has acted as a bit of a lag mechanism for demographic consequences to fully flow from the State to the Fed sphere.
David Richards: Ain’t that the truth? The Libs deserve to be in Ran until the end of time. They have learned nothing from their experience. Oz needs an opposition. Give the fools meth, that would energize them.
In my dreams, nightmares and day-dreams swirl and dance in patches of fog. One night I dreamed of a political party being elected in the normal way, BUT with a fundamental difference. Each Front Bench member of the Party would have one three year stint as Prime Minister. ONE stint ONLY. Pay Parliamentarians a lot more money. But cut their severance pay in half.
This would help to ameliorate the savage desire for power and self- aggrandizement of people like Rudd and John Howard.
What it would do for the country is problematical. Just a dream.
I question the reliability of this study. Aaron Martin was my tutor last year for political science and is very conservative, that similar to Andrew Bolt. Aaron tend to stretch the truth a fair bit, similar to Andrew Bolt. I understand that Dr. Juliet Pietsch is also a conservative and I question whether this “research” is a ploy to get swing voters to vote for the Coalition, given that it is not looking good. That said, should be a good read.
Penny, what a lot of people don’t appreciate (and I’m not getting into you here) is that when it comes to the pointy end of academic research, the overwhelming majority of researchers follow the numbers they produce from the empirical analysis they undertake regardless of whatever their personal opinion may happen to be about the subject.
And we aren’t really talking overwhelming like 75% or so – we’re talking overwhelming as in 99%+. “Making stuff up”, or even bending the results to suit ones personal views, is the quickest way to completely destroy your career. Once upon a time these things could be hushed up if they occurred – but not anymore.
I’d argue that the most trustworthy results come from academics that make an empirical based finding that runs contrary to their personal opinions, or how they’d like the world to be working if they could magically transform it.
We don’t have to agree with the politics or otherwise of a researcher (and some of the best intellectual shitfights we’ll ever have in our lives are with our learned betters at Uni) – but we should treat solid empirical findings like this on the basis of their own merits.
Otherwise, we just end up getting caught in personalities and partisan politics – and we can easily miss the quality contributions to the sum total of human knowledge that often come from places we ordinarily might not expect.
Possum Comitatus, I completely agree with you, and I know that other academic researchers, conservative or not, would be based on actual findings. however, knowing Aaron for that painful tutorial last year, it really wouldn’t surprise me if the findings were a little bent.
Possum@30 – I thought 2004 was more about interest rates. It comes down to another difference between state and federal levels – control over the national economy. Hence the push by the Liberals as better economic managers, a mantle they seem to have now blown.