12 Sep 2013

All Those 'Undeserving' New Senators

By Cara MacNish
Likely incoming WA senator Glenn Lazarus
Likely incoming WA senator Glenn Lazarus

Have the new minor party senators won seats they don't deserve? Actually, writes mathematician Cara MacNish, the preferential system has done its job

Media commentators have lined up since Saturday night to pour scorn on a system that has seen candidates with "less than 1 per cent of the vote" potentially elected to the Senate.

While the apparent irony makes for an enticing story, the implication that the candidates are undeserving of their seats based on first-preference results incorrectly erodes confidence in the electoral system. There is nothing in the preferential voting system used in the senate that demands that there is, or should be, any particular reverence afforded to the number one spot as a thermometer of public opinion.

On the contrary, the system is designed to reflect accumulated weight of opinion across the ballot paper.

The “undeserving” argument rests on the assumption that the success of the minor party candidates (notable exceptions such as the Liberal Democrats aside) lies with preference harvesting and cynical back room deals.

But there is an alternative explanation — that the results do in fact mirror public opinion, and that the specific minor parties chosen didn't bubble randomly to the surface from the large pool of sometime quirky contenders, but rather are those that the public as a whole are relatively comfortable with. In other words, the preferential system has done its job.

Every voter who on Saturday battled with the long white scroll below the line will be familiar with the feeling that, after the first dozen or two candidates, one is not so much choosing which platforms one subscribes to the most, but which platforms one finds the least objectionable.

This is natural in a democracy that supports diverse and opposing views — effectively you are not ranking candidates in the order in which you would like to have them installed, but rather voting for some and against others, with varying degrees of conviction. A more descriptive system would assign zero to the candidate you are most neutral about, positive numbers to those you wish to win, and negative numbers to those you don't.

As more polarising candidates fall away in the tally room, the votes of constituents who ranked them most highly give way to votes for candidates those constituents can live with. These candidates gain support irrespective of their first-choice status. A positive way to look at this is simply as a formal system for reaching a compromise on a large scale.

While the data are yet to be revealed, this is a plausible explanation for the outcome in my own state, where a candidate running on what may be perceived as a relatively benign ticket promoting sport and healthy lifestyles may be destined for success.

Why, then, is the effect more noticeable in this election? Have the preference touts been so much smarter, or the back room deals craftier? Perhaps, but again there is an alternative explanation.

The macro system of cumulative support reflects the micro system of the individual voter with the long white scroll. While the preferential system is more complex than adding voters' positive and negative leanings, there will be a gradual shift in the origin of support going down the rankings. One can imagine a "point of inflection", above which support comes primarily from first and high-ranking preferences, and below which it comes primarily from the more benign alternatives chosen by voters who's first choices have been eliminated. Put simply, from the "most wanted" to the "least unwanted".

Where this inflection occurs in the raking will depend on macro issues, such as the mood of the voting public.

In an election year where there is strong confidence in two or three major parties, the point of inflection will likely be forced down below the top six candidates, and the senate will be dominated by the major parties — a more "traditional" senate. Where there is large dissatisfaction with major parties, the point of inflection will rise — not because those bubbling into contention are most liked by the populace, but because they are least disliked.

Of course those voting above the line hand over responsibility for "most liked" and "least disliked" to a third party. The above analysis does not preclude the influence of preference harvesters, though their influence might not be as great as they would like to believe. Nor does it preclude parties distributing their preferences in contrast to their stated leanings or ideologies, perhaps surprising their above the line voters. However it does suggest that neither of these are necessary for the outcome we are seeing, and that the senate results may provide a better reflection of underlying voter sentiment than it first appears.

Log in or register to post comments

Discuss this article

To control your subscriptions to discussions you participate in go to your Account Settings preferences and click the Subscriptions tab.

Enter your comments here

Posted Thursday, September 12, 2013 - 12:26

There is nothing in the preferential voting system used in the senate that demands that there is, or should be, any particular reverence afforded to the number one spot as a thermometer of public opinion.

Yes.  That's the problem that we're all complaining about.  We're not saying that the system didn't follow the rules of the game; we're saying that the system is borked.


A dairy farmer goes to her vat expecting to see creamy white milk.  Instead, she sees that the vat is full of urine.  Extremely puzzled by this, she calls the company who built her milking machine.


How pleased do you think our hard-working, salt of the earth dairy farmer will be when the consultant says: 'Your expectation for milk instead of urine incorrectly erodes confidence in the milking system. There is nothing in the milking system used on your farm that demands that there is, or should be, any particular reverence afforded to milk rather than to urine.'


I think the dairy farmer would expect some explanation for why a tube goes anywhere near the urinary tract in the milking system.


And thus we return to the senate selection system.  Why is there a tube that goes anywhere near the tract of candidates who polled ridiculously low first preference votes?  How could a sensible system result in the situation in South Australia where Senator Hanson-Young's election resulted in votes flowing to Family First's Bob Day?  How could it be that Nick Xenophon -- who polled over a quarter of the first preferences in his state -- only ends up with one seat?


We should be able to test whether there was dissatisfaction with the major parties.  Let's look at Western Australia.  70% of the population voted for either the Coalition or the ALP.  Meanwhile, the microparties as a complete group, managed only to snag about 14% of the vote -- and that's including the 'Crap, we thought we were voting for the Liberal Party' vote from the LibDems.


There is no way that the result is representative of the state's voting intent.  Until something is added to the system (such as excluding on the first count everybody who polled fewer than one in ten first preference votes), the senate will continue to throw up nonsensical outcomes.

Posted Thursday, September 12, 2013 - 13:00

Lazarus is from QLD, not WA.

John Salmond
Posted Thursday, September 12, 2013 - 15:31

An interesting story. It provides a useful base from which to move on to more high-level political considerations about the voting system.

markfletcher's comparison is too extreme, though his core point was worth making (more briefly)

I think an important point is made about the 3 major parties all being seen as less desirable than normal.

One might suggest that it is precisely in such a situation that the wild card of independents are  usefuly brought into the equation. Naturally, the big boys will be trying to play the independents - but that is parliamentary politics

Posted Thursday, September 12, 2013 - 15:57

The obvious flaw in the system is that it didn't give the Coalition control of the Senate.

Posted Thursday, September 12, 2013 - 16:13

Here's some suggestions for reform.



Posted Thursday, September 12, 2013 - 17:22

If you have a quota of first-preference votes below which Senate candidates are eliminated, you need to set it low enough that the candidates who aren't on the top row aren't all eliminated once the first preference votes are counted. The existing quota is probably the right level.

What people dislike about this election is not the number of votes received by the winning candidates but rather the parties with which some of the winning candidates identify. If you want a quota to stop this from happening in future, you need to base it on parties rather than first-preference votes for candidates.

Maybe, you could set a quota on above-the-line votes but that would disadvantage a party that appealed to voters who valued choosing their preferences over the convenience of above-the-line voting.

This user is a New Matilda supporter. ErikH
Posted Friday, September 13, 2013 - 21:02

The best solution is to allow preferential voting above the line.

Mind you, the whole thing is compromised, as I undeerstand it, by the fact that preference distribution in teh Senate is by "sampling", not by "population". In other words, they don't distribute preferences as in the House of Reps (eliminating one party after each count and distributing its preferences) but rather, taking samples of the "rejected" votes and eliminating parties and distributing preferences according to those samples. 

Although that's kind of understandable given the numbers of votes and the complexity of the Senate voting system, it still seems very suspect.

The sooner we have electronic voting and can get rid of the sampling, the better.

PS I'm happy to be corected on the Senate preference distribution method - that is the way it was explained to me some time ago.

This user is a New Matilda supporter. ErikH
Posted Friday, September 13, 2013 - 21:20

Also on all the typos. Sorry

This user is a New Matilda supporter. RossC
Posted Sunday, September 15, 2013 - 17:44

I suggest you stick to mathematics in future Cara, because the 'subtleties' of politics - like for example, the candidate with the most primary and early preference votes deserves to win - are obviously lost on you.

Someone who 'wins' by a fortunate cascade of meaningless preference deals down to spot 100, but not because anyone actually liked them - or their professed causes - enough to give them a meaningful primary vote, is farcical.

Who are these drongo's actually representing? 0.1% of the primary vote, that's who - ie them and a few of their idiot mates. Which mathematics school did you attend that suggests that could be, in any way, representative, fair or balanced?

No statistical argument can convince me this is anything but a sad joke, and I think a lot of other Australians feel the same way. (the only ray of light I can see is that Tony Abbott will have to deal with these clowns to get stuff through the senate. That might be hilarious (and Tony deserves no better).

Posted Monday, September 16, 2013 - 09:39

Good article - I agree with the key descriptive of the Senate system as  "a formal system for reaching a compromise on a large scale".

Further, the mini-parties won according to the rules.

The mini "Sport Party " may seem ridiculous to some but it at least promotes a healthy lifestyle in contrast to the pro-war, pro-Zionist, war criminal, pro-coal, pro-gas, anti-environment, anti-science, climate criminal, US lackey, human rights-abusing major Lib-Labs (Liberal-Laborals, Coalition-Labor) under whose anti-Australian policies an estimated 66,000 Australians die preventably each year  (see Gideon Polya, “Why PM Julia Gillard must go: 66,000 preventable Australian deaths annually”, Countercurrents, 21 February 2012: http://www.countercurrents.org/polya210212.htm  ). Indeed it is estimated that the budgeted cost plus longterm accrual cost to Australia of the Zionist-promoted War on Terror is $125 billion, this being associated with 12 years  x 66,000 preventable deaths  per year = 792,000 preventable deaths since 9-11 (see Gideon Polya, “Endless War on Terror. Huge cost for Australia & America”, MWC News, 14 October 2012: http://mwcnews.net/focus/analysis/22149-endless-war-on-terror.html ).

Of course the real Elephant in the Room (Senate Chamber)  is that Tasmania (population 0.5 million) has the same number of senators as Viictoria (population 5.6 million ) or NSW (population 7.4 million). As former PM Paul Keating said, "Unrepresentative swill" and one could further well query  how many votes US citizen Rupert Murdoch has in Murdochracy, Lobbyocracy and Corporatocracy Australia in which Big Money buys people, politicians, parties, policies, public perception of reality, votes and political [power (see David Willamson's latest play "Rupert" and see "Boycott Murdoch Media" https://sites.google.com/site/boycottmurdochmedia/  ).

That said, the proportional representation element of the Senate has meant that the decent, pro-peace, pro-science, pro-environnment,  anti-racism, pro-human rights, and progressive Greens have had a very useful public voice in trying to save Humanity and the Biosphere  from the racist, grossly human rights-abusing, pro-war, pro-Zionist, US lackey, war criminal, anti-science, climate criminal, Gadarene major party Lib-Labs.