Wednesday, April 3, 2019

Fraser Anning, How-To-Vote Cards And Bad Electoral Reform Proposals

Senator Fraser Anning has not endeared himself to most of Australian politics in his short Senate career.  Elected on a special count as a replacement for One Nation's Malcolm Roberts, he left One Nation amid mutual distrust very soon after his arrival.  He then joined Katter's Australian Party before being kicked out of that for being too extreme.  His scorecard includes use of the term "final solution" in his maiden speech, blaming the Christchurch massacre on the fact that Muslims were allowed to migrate to New Zealand (huh?) and proudly attending neo-Nazi rallies.

I've seen a lot of discussion about Anning and how he got into the Senate, and I've noticed two trends that I think need to be dealt with.  One is that a commentator, hellbent on preventing any future Annings from getting into the Senate, comes up with an electoral or parliamentary reform proposal that would either have prevented Anning getting into the Senate the way he did, or else at least allowed for his expulsion as soon as he got there.  Then, because they've found something that would get rid of future Annings, they promote this idea without thinking through any greater problems it might cause.  The second, though I haven't seen so much of it just yet, is that a commentator who already has some electoral reform proposal they want to support, looks for some way they can argue for it by making a point about Fraser Anning.

I've been meaning to make some of these points for quite a while, which is why this piece is so long, but there were two catalysts for getting it done today.  One was having our electoral system criticised by Winston Peters, who not only falsely claimed that Anning is there by "pure accident", but also suggested we needed to change our electoral system to stop such people being elected in the future.  I previously had some silver ferns on this site's background in sympathy over what happened in Christchurch at the hands of one of this country's many home-grown fascist losers.  However, there is no excuse imaginable (not even if it prevented the end of the world) for an NZ government MP to lecture Australians about our electoral system, so I have had to take those down.

We Will Not Be Lectured By Anyone From Over The Ditch

New Zealand has a multi-member proportional system that combines the abomination of first past the post and the crassness of primary vote threshholds for proportional representation.  Moreover, the use of single-member electorates enables a party to rort around proportionality by throwing seats to a flank party - by this method something called "United Future" won a critical balance of power seat off 0.22% of the national vote in 2014.  Some people argue that New Zealand's MMP system is not a bad hybrid system to have if you are going to have a single house of parliament, but that presumes that you would want to do something so foolish in the first place. So I think before any New Zealander lectures us on how we elect our Senate, he ought to enlighten us about how they elect theirs.

NZ is also rife with exactly the sort of party-hopping ("waka-jumping") that has led to Fraser Anning being free of whatever questionable restraint parliamentary membership of One Nation might (or might very well have not) have had on his recent behaviour.  After decades of it, NZ has implemented an extreme solution, which would be a very bad idea if applied here.  More on that later.

Anning's 19 Votes

One source of objection to Anning being a Senator is that he polled a risibly small number of primary votes in the 2016 Senate election (to be precise, nineteen).  This echoes a similar criticism also made of the election of Malcolm Roberts, who polled just 77 primaries on his way to winning in the original election.  Roberts was later disqualified and the special count without him elected Anning.  

Senators can be elected with very small primary votes in Australia because we have such things as preferences - ours is not some lame first past-the-post system where nothing but a number 1 vote matters.  Sure, Anning only got 19 primary votes (actually in the special count he picked up nine of Roberts' and commenced with 28).  However he was powered to victory by:

 (i) the fact that Pauline Hanson had more votes than she needed to be elected, and therefore was entitled to transfer spare votes to other candidates (otherwise, a party would be hurt if its vote was concentrated in its most popular candidate)

(ii) the fact that many other Queensland voters, having not voted for One Nation, preferenced them in droves, helping them to swamp other smaller parties in the preference count, overtaking six other parties in the process.

It's not really clear what reform proposal those complaining about the 19 votes thing have in mind. But if someone says that it shouldn't have been possible for Roberts/Anning to win off less than 100 primaries, then they should also apply that to many other Senators (see some examples here) elected off the back of party votes of a few hundred down the ages.   

There are electoral reforms that could be pursued to make this much less likely to occur (for instance: remove above the line voting, rotate candidate names within each party's list), but I have never seen anybody who proposed them even try to cost the increased budget for AEC data entry that would result from doing so.  And there are other potential hazards with such an arrangement, including whether big parties might unfairly benefit from the ability to spread the vote across individual candidates.   But ultimately, if a party's voters insisted on all putting a given candidate first and another given candidate second, and the party had enough for two seats, why on earth shouldn't the second candidate be entitled to win?  

(Even in Tasmanian Hare-Clark, which has the above features with how-to-vote cards banned on the day as well, one sometimes sees very lopsided results within a party.  For instance Judy Jackson (Labor, Denison) was elected 5th of 5 seats in 2002 despite running tenth on primaries with only 2.6% of the vote.  Her election was thanks to the huge flow of preferences from party colleagues.)

Anning's Party-Hopping

The next bad reform proposal I've seen a lot of is one that argues that, yes, Anning was elected fair and square on the special count, but having been thus elected he should have had to sit as a One Nation Senator, and not been able to go off on his own, for which he had no mandate.  So the proposal is that on leaving One Nation, he should have forfeited his seat.

However being a member of a parliamentary party places no restraints on a Senator's behaviour whatsoever if the Senator doesn't want it to.  Anning could have crossed the floor, behaved as he has done, ignored other One Nation Senators and so on until they kicked him out.  Either the forfeit proposal relates only to Senators who quit their party or it includes those who have been expelled.  If the former, then it is no restraint at all, since a Senator can act the goat in whatever manner they will while refusing to resign from their party.  If the latter, then if a party can cause a Senator to be kicked out of the Senate by expelling them from the party, then that gives parties too much power over their own Senators.  For instance, Labor could have responded to Lisa Singh's below-the-line victory by expelling her from the party and hence from the Senate, and then replaced her (as a casual vacancy) with the candidate they preselected ahead of her.

Indeed, this is exactly the problem I have with what has been passed in New Zealand.  If an NZ MP is kicked out of their parliamentary party and 2/3 of the members of that PP want that MP expelled, then that MP can be expelled and replaced with the next one on the party list (unless they hold an electorate-based seat, in which case there's a by-election).  So Labor could have expelled Singh, the Greens could have expelled Lee Rhiannon (over internal disputes that the recent NSW election showed to my surprise that virtually nobody cares about) and so on.  A milder version might be to apply the expulsion power specifically to Senators who come in on special counts, but that would mean any party who lost a Senator to Section 44 could just expel the replacement and reinstate the original Senator (if now eligible and that Senator's state approved, of course).  And it might be that the party had gained an unfair advantage by running that ineligible person as a candidate to begin with.

Expel Anning For Being Anning?

The next version is that Anning should be expelled specifically for the nature of his extreme and crackpot statements, and not for the manner of him getting into the Senate, nor of him leaving One Nation.  This would, firstly, require both houses of parliament to repeal Section 8 of the Parliamentary Privileges Act 1987 (which rescinded the expulsion power inherited from the UK according to the Constitution).  I have seen some informed legal minds express the view - but not in any considered or formal form - that the High Court might get picky about whether bringing back expulsions would implicitly contradict "elected by the people" (and other such concerns), but leaving that aside, bringing back the expulsion power is risky.  Admittedly, expulsion threats at state level have recently proved successful in hastening the resignations of disgraced MPs involved in corruption or resume fraud, but the Constitution already provides for the expulsion of MPs convicted of serious offences or with certain conflicts of interest.  The expulsion of Hugh Mahon in 1920 was grossly unreasonable and it would be better to leave any future Annings to face the judgement of voters at the next election, than to risk an MP being expelled by the party that had the numbers to do so, for a reason unworthy of expulsion.

There are also issues with the nexus between expulsions and casual vacancies.  An expulsion creates a casual vacancy filled by a joint sitting of state parliament which can only nominate a member of the party that originally won the vacant seat (if that party still exists).  So there is nothing to stop the original party from renominating the expelled member or else nominating somebody even worse, if it wanted to.  (In this case, Anning's seat would be filled by One Nation).  We could even get a situation in which someone like Anning (but not a party-hopper) was repeatedly expelled, reinstated, expelled, reinstated and so on.

Fix Section 44 So There Are No More Annings?

This argument actually isn't as bad as the other ones, and fixing S44 isn't actually a bad proposal (on the contrary).  The Anning-specific argument is that had Section 44 not been so ridiculous in the area of meaningless joint citizenships, Roberts wouldn't have been booted, and Anning wouldn't be in the Senate to begin with.  Moreover, it can be argued that Anning's extremist behaviour is a consequence of One Nation kicking him out - he's bound to try to get votes for himself somehow.  We have now seen eleven Senators kicked out by Section 44 in Australian political history, with four of these cases (Robert Wood replaced by Irina Dunn, Malcolm Roberts replaced by Fraser Anning, Jacqui Lambie replaced by Steve Martin and Skye Kakoschke-Moore replaced by Tim Storer) resulting in the replacement MP leaving the party they were elected under.  That's a high ratio, and it arises mainly from parties putting unreasonable pressure on the replacement to quit so the original Senator can be reinstated.  So, reduce Section 44 disqualifications and you have less risk of replacement Senators zooming around without mandates.

There's a very good case for fixing Section 44 without all of this.  It fails to correctly capture the essence of inappropriate loyalty (Sam Dastyari was doing the bidding of Chinese interests, but had gone out of his way to rid himself of all dual citizenships - yet most of the Section 44 victims had no idea of their allegiance and wouldn't have been motivated by it if they had).  It creates constant legal uncertainty, and it encourages parties to be - well, racist basically - in deciding which candidates can safely be preselected.  But I'm not sure Anning adds much to the case.

Firstly, it's not so clear that Anning had no mandate*.  Voters for One Nation willingly voted for a party with a history of being a total rabble in the state where Anning was elected.  They did it knowing One Nation expressed many Anning-like views (though not as extremely) and with ample opportunity to easily research Anning's views if it concerned them he might win.  He has more of a mandate from the careless or deliberate consent of One Nation voters than Senators elected in the old days of Group Ticket Voting, in which it was impossible for voters to have any real idea who their above the line vote might help elect, and unreasonable to force them to number every box if they wanted to avoid their preference electing an Anning type in any case.   

And secondly, if we're going to point to Anning's election as an argument for Section 44 being bad, let's look at all the good it's done for politics!  It removed Malcolm Roberts, who appears to be a much nicer person than Fraser Anning, but nonetheless had been the Senate's most platformed climate change denier, and hence arguably far more dangerous.  It evicted David Feeney, one of Labor's worst of faction hacks, and replaced him with Ged Kearney who is much better placed to defend Labor voters' concerns.  It brought diversity to the Senate in the form of Jordon Steele-John, who may not be Scott Ludlam and does appear to swear a bit but who successfully pushed for a royal commission into abuse of people with disabilities that I do not think would have happened without him.  If you're concerned about Islamophobia, then Jacqui Lambie has had form in that regard in the past - S44 got rid of her as well.  It reversed a grievous preselection wrong by restoring Richard Colbeck to the Senate.  And so on! So judging Section 44 by whether it brings us good or bad politicians is all very subjective, and shouldn't really be the argument at all.

(* Looking for evidence on what voters could have known about Anning prior to voting if they wanted to, I found one of the very few pieces that seems to have looked at Anning's views prior to the 2016, and that is this piece by Luke Henriques-Gomes regarding Anning's then-offline old Facebook account.  I was surprised to find when I clicked on the link to "Mr Anning’s Facebook page" in that article that the link had again become live.  It contains much of the material referred to in the article, but also contains shares of extremist anti-Islamic material by the Patriots Defence League - Australia, a rebadge of the Australian Defence League, and a group which overlaps with Blair Cottrell's United Patriots Front.  

A PDLA meme shared by Anning on Oct 21 2015 and again on Jan 7 2016 and which I decline to link to or reproduce alleges that "The Path Of Islam Is Always The Same" and refers to a supposed pattern of takeover through enclave formation, rapid reproduction, resistance, Sharia law and secession.  Hilariously it claims "Pakistan was once Hindu" as an example of this pattern.  Most likely if One Nation's vetting even detected that Anning was approvingly sharing such material, the party wouldn't have cared anyway.  Either way, the warning signs that Anning could take an extremist position on Islam if somehow elected were there.)

How To Vote Cards Caused Anning?  What?

Last and worst of all, on to one of the worst op eds on electoral reform that I've seen published lately.  (I've Waybacked it because some behaviour should not be rewarded with actual link clicks.) It's a proposal for banning how to vote cards.  It's the sort of thing you get when newspapers accept op eds from someone with no greater stated qualifications than "freelance writer" and then figure that since it's only "opinion", it needn't be edited for errors of fact - or for simply being totally illogical.

To start with what is wrong with the non-Anning-related content of the article (since I think that should be mentioned too) it argues that allowing how-to-vote cards advantages big parties since big parties have more resources to hand them out.  It presents no evidence that big parties actually gain votes by handing out how-to-votes.  If this was true, one would expect the state where how-to-votes are banned (Tasmania) to have high micro-party vote rates at its state elections, but it doesn't (over 93% of Tasmanians voted Liberal, Labor or Green at our 2018 state election).

The article alleges that voters are "pawns in a game decided through preference whispers and backroom deals".  It ignores the empirical evidence from countless elections that only major parties have high HTV card follow rates - which means that how to vote cards seldom decide seats, because major parties are seldom excluded from the count.  The large majority of voters for non-major parties make up their own minds.

It argues that "Also, compared with other Western democracies, Australia is unique in its heavy reliance on how-to-vote cards." and attributes this to compulsory voting.  Wrong.  It is attributable to preferencing.  Obviously, countries with non-preferential systems, which is nearly all of them, do not need how to vote cards.

Most importantly, the article completely misses the major point of how-to-vote cards.  The point is to assist voters to vote formally.  It's all very well to talk about how engaged, informed, English-speaking citizens can vote without the things (as Tasmanians do at state elections), but in areas of low literacy and/or low English-speaking rates, the cards are necessary because without being able to copy them, otherwise loyal voters for a party will ruin their vote.  Indeed, these areas have too many informal votes as it is, without taking away HTV cards and making it worse.

There's probably even more wrong with this piece, but it's time to get onto Anning.  This is what the article says about Anning:

"We can also thank how-to-vote cards and preference deals for the unwelcome presence of politicians who few people actually vote for. Fraser Anning, anyone? If more people were encouraged to understand the benefit of casting their own preferences and voting below the line in the Senate, we would avoid the fringe dwellers."

Not only is this false, the very opposite is true!

Anning was elected in the special count off One Nation above the line votes, below the line votes for Pauline Hanson and other PHON candidates, and preferences from other parties.

To start with One Nation, they had 250,131 votes in the special count.  Of these, just 11,980 followed the how-to-votes, meaning that Anning gained just 1943 votes from ON how-to-vote cards via Hanson's surplus.  Anning's victory margin was several times this.

Anning's final opponent in the Senate count was Family First.  Only six parties are known to have preferenced One Nation on their how to vote cards at all.  These were Australian Liberty Alliance at 2, Shooters Fishers and Farmers at 3, Rise Up Australia at 3, Katters Australian Party at 6, Family First at 6, Democratic Labour Party at 11.  Family First was not excluded until One Nation's election was assured, and KAP, DLP and RUA all preferenced One Nation above FF.  So the only parties whose HTV card preferences as shown by the ABC reached One Nation at all were Australian Liberty Alliance and Shooters, Fishers and Farmers.  The ALA card was followed 1-6 by 4761 voters  and the Shooters card by even fewer people than voted 1 for Anning (17).  So even had no preference that reached Anning via a how-to-vote card done so, he would still have won, unless all the preferences pooled very strongly with his main opponent.

But more, One Nation received immense piles of preferences that went against the how to vote cards. From above the line #2s alone, Anning received at least 28,274 votes by voter choice for One Nation from parties that either didn't preference One Nation at all or else preferenced them below Family First.  There were tens of thousands more further down the ballot paper.  This included One Nation being the most popular destination for voters for parties such as the Lazarus Team, which didn't even preference them at all!  One Nation's win of two seats had nothing to do with party preference deals and how-to-vote cards and was achieved not because of how-to-vote cards but in spite of them, at an election at which micro-party voters rejected how to vote cards totally - exactly what this freelance writer person wants.

More may be added if any more silly Anning-related arguments are seen.  


  1. Would changing the counting method to count votes for parties as equal rank for all the party's candidates do much?

    1. It would certainly stop people winning with 19 (or 28) primary votes to their name, but only through an artificial method of deeming party votes to be spread across the party's candidates when (given the choice) the party and all its candidates might prefer otherwise.

      It would have some other, and quite major, impacts. One of them is that it would help major parties to spread their votes evenly across their candidates, which (in a system where votes exhaust) would advantage major parties. For instance, if a major party has 2.4 quotas and is fighting for a third seat against a minor party with 0.6 quotas, the major party will probably win three seats and the minor party none, whereas under the current system it would probably be 2-1.

      It would also mean that if a party has a high above the line voting rate, which of its candidates wins is determined by the small number of below-the-line voters for different candidates for a party (not only for that party but also preferences coming in from other parties). This could be deliberately engineered - eg a right-wing religious party doesn't like a left-wing Labor Senator so it convinces its voters to vote below the line and direct preferences to a more right-wing Labor candidate. If all voters started voting below the line this would become more like Tasmanian Hare-Clark, but until they did it could produce some weird results.

  2. Hi Kevin,

    Thankyou for an excellent and informative article.


  3. What about a supermajority (or more) of Senators, or a petition of enough signatures, being able to force a state byelection? It would be a very rarely used power, considering that it would be expensive and unpopular, but it would be a "break glass in case of emergency" option that would never override the will of the people.

    1. A state by-election would mean that if a minor party Senator was ousted they would get replaced by a major party one. Unless it was a by-election for the whole slate as in WA 2014, which would be a bit harsh on the rest and likely to produce distorted results (as WA 2014 did with high voting for Greens and PUP).

      I do think there should be something stronger than censure for MPs who severely disgrace the Parliament with their behaviour outside it. A three-quarters majority for a suspension of modest duration, perhaps.

  4. Suspension for a time I agree with but removal would be a step to far. Imo disqualification would open the bottle that let the genie out and I'm not sure it could be put back. Imagine a government that already had a majority in the senate, they could use this power to lessen opposition even further.

  5. Should of added, when it comes to HTV cards I have no idea. I can see their benefit for people with poor literacy in the english language. So how do you devise a system that allows those people to cast a valid vote without pointing them to party x, y, or z when they may prefer the candidate from party a?

    1. It's very difficult to deal with the problem of high unintended informal voting in a compulsory preferencing system. One option is banning how to vote cards but switching to optional preferencing, but this would be highly divisive because it would be perceived at various times to advantage one party or other. In the past I've supported OPV on personal freedom grounds but I am cooling on this because I am sick of seeing parties put out deceptive "Important: Just Vote 1" style messages, which in my view under a properly working OPV system would need to be very harshly punished to avoid electors being deceived.

      Some progress can be made by improving savings provisions. Antony Green has suggested regarding any vote that only skips numbers as formal (treat 1-2-4-5-6 as 1-2-3-4-5 etc). I support this and think there are also ways to save votes with doubled numbers while preserving full preferencing (by randomly re-ordering the doubled numbers). But these changes can only go so far.

    2. Why not a standard 'How to vote' card issued by the AEC, and then that's the only thing anyone can hand out near the polling places?

  6. That article about banning HTVs had all sorts of nonsense in it, but I do have a problem with them. Saying "This is 'how to vote' Labor/Liberal/etc" carries a strong implication to the unsophisticated that it is the ONLY way to do it. If the things are not to be outlawed, the use of the phrase "how to vote" should be. It should be compulsory to point out that the preference recommendations are only recommendations, and that the vote will be valid if the voter numbers all squares in any order at all. (Amazingly, among the HTVs that I have seen, the ones that come closest to doing that are from One Nation. Fancy them being the most honest party!) Likewise, any "Just vote 1" propaganda should be qualified with a reminder (not in small print) that it's only a recommendation. As I recall it, Beattie's original ones were styled to look like an official ECQ poster, with just a tiny "Authorised by" at the bottom. Effective, but disgracefully sneaky!

    1. And pretty much the same tactic as Beattie's was seen in NSW recently - the signs put out by the Nationals and in some places by Labor looked like official material though they did have party logos on them if one looked closely.

      Yes I would support greater restrictions on the wording of how-to-vote cards (and also those handing them out) so that it must be clear to voters that the order is only a suggestion and that their vote will be formal provided all squares are numbered without omitting or repeating any numbers. Perhaps they should be renamed "Voting Suggestion Cards".

  7. On the following of HTVs:

    Back in the days when The Wilderness Society was running the first of what are now called "Third Party Campaigns" (which ultimately preferenced the ALP over the Coalition), I did one or two trials in electorates in which we issued different HTVs (with slight variations between candidates for whom we had no particular preference) at adjacent booths in the same electorate.

    Scrutineering showed that a very consistent 2.3% followed these at the booths where they were issued.

    GetUp! has started doing this stuff - did so in Wentworth, will do so here in Warringah. I will be interesting to see what happens. This behaviour, I think is very much influence by the number of candidates - certainly with Optional Preferencing.