Last week the Australian Electoral Commission had an interesting media release about the estimated 1.2 million eligible voters that aren’t on the electoral roll.The media release stated:
In the continuing search for 1.2 million Australians missing from the electoral roll, the Australian Electoral Commission (AEC) released information today indicating just who in Australia has the greatest chance of risking their vote in the next election. Electoral Commissioner, Ed Killesteyn explained that Australians aged 18 to 34 years or anyone who has moved house in the last three years, have the greatest likelihood of not being on the electoral roll. “The fact is the younger you are the less likely you are to be on the electoral roll, with young people aged 18-24 years accounting for nearly one third of the 1.2 million eligible Australians missing from the electoral roll,” he said.
So that got me thinking, what would have happened in the 2007 election if those missing from the electoral roll actually enrolled, turned up and voted? Would it have changed the election result – if so, by how much?
First up, we need to look at the AEC estimates of the age breakdowns of this missing 1.2 million odd people.
As we can see, the younger you are the more likely it is that you aren’t on the electoral roll. To show the approximate difference between the actual electoral roll at the 2007 election, and a hypothetical electoral roll where these 1.2 million eligible voters are added, we can take those above numbers as given for their minimum values. For example, we will assume that rather than there being over 370,000 18-24 year olds missing, we’ll assume that there actually is 370,000 18-24 year olds missing. This way, we’ll end up with a conservative estimate – a “if all eligible Australians voted at the election, it would have changed the result by at least this much” type of thing.
Once we combine these numbers with that of the electoral roll in 2007 we get:
This shows what proportion of the electoral roll each age cohort makes up, both at the 2007 election and for our hypothetical “full enrolment” alternative. What we see here is that young people would make up a larger proportion of the voting electorate with our hypothetical “full” electoral roll compared to what we actually had in 2007. 18-24 year olds would be 1.6% larger in weight, 25-34 year olds 0.8% larger – while the older cohorts would make up a smaller weight in voting terms were we to have a a 100% enrolment rate.
This is important, because we know that the younger you are, the more likely you were to vote for the ALP in two party preferred terms in 2007. So would this new age composition of the electoral roll have changed the election result, and if so, by how much?
To start things rolling on the vote projections, we need some good age breakdowns of the two party preferred vote at the 2007 election. Fortuitously, we have arguably the most accurate estimates in the country, by age, of the ALP TPP vote at the last election. You can find the gory details of how it was created over at the top of the Gen Blue post where we first used it (and it’s worth a read if you missed it).
Next up, we need to make an assumption that those people eligible to be on the electoral roll but aren’t, would have voted for the ALP at the same rate, by age cohort, as those who did participate in the 2007 election. That’s not a particularly heroic assumption – but it’s an assumption none the less, so keep it in your thought orbit. I’d be interested to hear anyone’s thoughts on this assumption – as I think those eligible but not enrolled would actually vote for the ALP in slightly higher proportions… but that still works for giving us a conservative estimate.
Once we use our two party preferred estimates by age cohort and apply them to both the 2007 electoral roll and our hypothetical “full” roll, we can determine the two party preferred contribution that each age cohort made to the final two party preferred result.
What we see here, is that the ALP two party preferred result was 52.7% at the 2007 election, but would have been at least 53.1% were the electoral roll “full”.
Where the change comes from is the points of two party preferred that each cohort contributes. In 2007 the 18-24’s contributed 6.9 points of the 52.7% final result, but under a full electoral roll scenario they were projected to provide at least 7.9 points – a full 1 point more. That comes about as their contribution is a function of both population size and the relatively high proportion of that cohort that voted ALP. Since their population would be much higher under a “full” roll scenario and they vote ALP at a higher level, they would end up adding more to the ALP two party preferred result under our hypothetical scenario then they did at the actual 2007 election.
You’ll also notice that the older cohorts provide a reduced contribution to the ALP TPP result – that comes about because those cohorts end up being a smaller proportion of the electoral roll once the 1.2 million missing people are added.
Under a hypothetical full electoral roll, the cohorts that vote strongest for the ALP become larger, while the cohorts where the ALP is weakest become smaller – resulting in an overall increase in the ALP two party preferred vote.
A 53.1% two party preferred result would have netted the ALP an additional 4 seats. Yet this is a conservative estimate, we are using minimum numbers here. The actual numbers are higher than we are using – we just don’t know by how much.
The ALP under Rudd is a brutal political outfit – they will wield every conceivable weapon to gain every possible vote at the next election. We should expect some serious voter enrolment programs to be unleashed between now and the next election – simply because of the fact that higher enrollments lead to higher ALP two party preferred vote share, both in terms of the 2007 result as well as the current polling.
In fact, if we use the current polling rather than the 2007 results, even though the ALP is getting a lower primary vote in the younger cohorts than it received at the 2007 election, so is the Coalition – in net terms their respective primary vote losses are nearly all going to the Greens, and flowing back to Labor via preferences. As a back of the envelope calculation, this pushes the potential gain to the ALP from enrolling voters out beyond a minimum 0.4% to well over half a percent of two party preferred.
No party sneezes at a free increase of their vote that’s at least up to 0.5% of TPP, especially when a 0.25% increase is easily attainable by spending some taxpayers cash on enrolment programs.
Expect to be inundated with them.
21 Comments
I found the most effective were the roving door-to-door folk – a few years ago they appeared at my then sharehouse and actually stood there filled in the forms with us.
Interesting point about the greens preference flowback. Do the percentage of prefs that return to the ALP change by cohort? As in would the 18-24 group preference the ALP in larger numbers when voting for the greens than other age groups? Or is it relatively stable?
Dan, I’ve heard from hacks that it’s a slight tilt towards the young, where young Greens prefs flow slightly more toward the ALP than older cohort Greens prefs – but I haven’t got or seen any definitive data on it.
Anyone else know or have anecdotes?
While i think the overall effect is probably quite true; younger people are less inclined to be enrolled, younger people preference Labor therefore a strong enrolment drive would help Labor… but i reckon there could be a slight (and it may be very slight) difference between the preferences of enrolled and unenrolled voters.
I’m thinking of people who deliberately aren’t enrolled as they don’t care or don’t want to be seen to support the system. Now i’m sure those passionately against voting may include some left-wing ‘conscientious objectors’, it would be my guess, and it’s only a guess, that more of them would come from a more right-wing perspective. Certainly those canny enough to not be enrolled as opposed to being enrolled and deliberately voting informally, i reckon, would include many who don’t want to pay the fine for not voting. Now who are the types that don’t want to pay fines to the government more than others? My guess is strong libertarian types.
I’m sure there are many other variables as well, and i may be wrong on that assumption, but it think it would be interesting even to know how many people aren’t enrolled because of a deliberate act rather than that they just didn’t get around to it or dropped off the role and never got back on.
Presumably the Green vote would have been higher with full enrolment. This might effect close Senate races and certain inner-city seats in future elections.
@thewetmail – I know of three people (30+yrs) who aren’t enrolled for that very reason, fear of getting fined. If the Gov was smart the should offer an amnesty and get them on the roll. I’m sure that all three would vote for either the ALP or Greens.
5
I believe that the Electoral Act specifies that enrolling on the electoral roll stops the AEC or Police from fining the non-enrolee.
@Tom the first and best (I love that tag btw
) Really? If that’s true I will tell them but I need to be sure. I will go have a look, thanks
Possum, great analysis, thank you. This should finally put to rest the ridiculous conspiracy theory thrown about after the election by some pitiful Liberal Senator (that’d be you, Senator Ronaldson) that the 1.2 million missing enrollees were the difference between the Coalition retaining power and an ALP victory.
In fact, they forestalled a much larger defeat with a shameful abuse of power.
I’ll add a series of slightly controversial assumptions. The rate of non-enrolment is almost certainly highly related to socio-economic status. e.g. Low SES people aren’t enrolled as much.
Capturing the non-enrolled population is most likely to boost the labor vote in safe seats and possibly cause minor boundary creep. Therefore it probably makes little electoral impact. However getting those extra voters in marginals would be worth the door-knocking.
EnergyPedant, Possum,
What EnergyPedant has contended rings fairly true. I wonder if, though, there may be a more significant impact in the Senate? Are there any States or Territories in which the addtional proportion of the 1.2million might make a difference to the last spot?
I would have thought that just by default the enrollment would increase next election. Under the old rules people could move within the same electorate and not have to re register. The new rule meant they had to re enroll and many people thought they were on electoral role only to be sadly mistaken. I assume that would account for some of the more middle groups in the data. AS those who were wrong on election day would now know to update their role it seems they would be more likely to get it right next time round. That is the impression i was under anyway.
An anecdote? An observation over a few blood-soaked years of handing out how-to-vote cards for a minor political party. The Liberal voters were far more likely to insult you than were the Labor voters. Actually, Labor votes were good at giving a sort of lop-sided, tolerant grin.
The drop from the 45-54 YOs @ 14% down to 7% for 55-64 YOs is harrowing. What is happening here? It’s too young for this group to be dying off, dementia? Diminution of civic duty, complacency? Perhaps the government should advertise the older groups as well as the young voters.
Venise – the Gen Blue article describes that big drop
http://blogs.crikey.com.au/pollytics/2009/05/15/gen-blue/
I’ll also have something to say about it tomorrow with an article about a new working paper that a couple of researchers are currently doing that covers that particular topic.
Another informative analysis Possum. Thank you.
Do you intend to analyse the 2009 Lowy Institute Poll? As far as I can find, we have had only media reports of the results, which seem somewhat to distort them, particularly regarding the contemporary debate on climate change. There has been much emphasis on the fact that ‘Tackling Climate Change’ has slipped from equal top to seventh place from 2007 to 2009 on the list of ‘Foreign policy goals’ (page 6), yet less has been made of the fact that 56% of those polled say it is ‘very important’ (fairly important, not very important or not at all important figures are not listed). Also on pages 27 and 28 there are more data about climate change that shows that 48% of respondents believe ‘Global warming is a serious and pressing problem. We should begin taking steps now even if this involves significant costs’, and in answer to the question ‘Would you say climate change is a problem or is not a problem?’, 76% say it is a problem. So although it might be true that concern among poll respondents about climate change has diminished since 2007, it still seems to be substantial. It would be a pity as the ETS debate hots up if the figures were used to give the impression that the climate change debate has lost its relevance to the public when there still seems to be a lot of concerned expressed in the poll figures.
What would help would be an incisive Possum analysis to reveal the true picture. The pdf file is accessible via http://www.lowyinstitute.org/Publication.asp?pid=1148 You may be able to get the data not revealed in the report, especially the ‘fairly important, not very important or not at all important’ figures for ‘Foreign policy goals’.
I’m sure your regular readers would welcome your appraisal of the Lowy poll results.
The flip side here is the danger that there is a trend where the younger cohort becomes less and less likely to register over time. And the ALP will actually go backwards.
Is there any data on that trend?
EnergyPedant – hardly controversial assumptions. people of low SES backgrounds, people of NESB’s, Indigenous Australians, Single mothers and people of low or no education, all of which would most likely be young people, are all groups that face the most disadvantage in Australia – it would make sense if they also experienced disadvantage when it comes to getting enrolled.
I suspect that conscientious objectors, Libertarians and people afraid of fines and prosecution make up only a small percentage of those unenrolled.
At first, I thought it might be interesting to do a projection based on the demographic of each electorate but, considering the level of precision, that would be to enculer les mouches (pardon my French).
Interesting analysis. As you suggest, the ALP will be putting a lot into this.
Thanks Poss, will go back to Gen Blue article, also look forward to your comment tomorrow.
Possum, has any allowance been made for Exclusive Brethren and other non-voters on religious, cult or sect grounds?
Steve – nah, they’re not large enough to matter.
I’m sure that Richard D’Natale(Green) would be a Senator for Victoria, and probably David Ristromm(Green) at the expense of Steve Fielding.
It’s about time the Electoral Commission showed a bit of independance and did these things.