Category: Electoral reform

As the Abbott government stalls on introducing the Senate electoral reform advocated by last year’s bipartisan finding of the Joint Standing Committee on Electoral Matters, The Australian today offers an opinion piece concerning the matter from Ross Fitzgerald. After extolling the vitues of Fiona Patten, the Sex Party figurehead who won election to Victoria’s Legislative Council in November, Fitzgerald takes aim at those who advocate reforms that might have prevented it:

So how did she get elected by a margin of some 20,000 votes? If you believe the nonsense spouted recently by psephologist Antony Green and Labor powerbroker Gary Gray, she “gamed” the proportional representation voting system under which the election was held. That is, Patten won a seat with a primary vote of about 3 per cent by cajoling and carousing with other minor parties to undermine a system that ideally should return “serious” parties like Labor, Nationals, Liberals and the Greens. These arguments need to be exposed before parliaments throughout Australia start acting on them in the mistaken belief that democracy is somehow better served by fewer rather than more political parties. Green maintains that minor parties, who tightly swap preferences among themselves and exclude the major parties, are somehow rorting the system. He and Gray appear to believe that the proportional representation system of voting only works properly when it returns a relatively small crossbench and preferably one that does not hold the balance of power.

Gary Gray I can’t speak for, but since Antony Green’s opinion on these matters is much the same as my own, I consider myself well placed to point out what is wrong with this.

Fitzgerald’s column is liberally sprinked with words like “apparently”, “appears to” and “as far as I know”, followed by assertions which in many cases prove to be untrue. Perhaps it does “appear” that Antony Green believes proportional representation “only works properly when it returns a relatively small crossbench” – but only if one is operating from a position of ignorance to which Fitzgerald, under the circumstances, is not entitled.

In fact, what Antony Green advocates is the model used for the New South Wales Legislative Council, a chamber of 42 members that includes nine cross-benchers who were elected as such. Among these are two members from Shooters & Fishers and two from the Christian Democratic Party, along with five Greens. This makes the situation hardly different from the Victorian Legislative Council, a chamber of 40 which includes five Greens along with another five from various minor concerns. Indeed, Fiona Patten would probably have been elected as lead candidate of the Sex Party under the New South Wales model, given her party’s 2.6% share of the Victorian statewide vote.

Fitzgerald continues:

As far as I know, neither of them levelled similar criticisms at the independent Tasmanian senator Brian Harradine, who held federal governments to ransom for nearly 20 years off a primary vote of 7 per cent or when Family First’s Steve Fielding was elected to the Senate off a primary vote of 1.8 per cent.

If Fitzgerald feels he can use the phrase “as far as I know”, and follow it with the assertion that Antony Green was relaxed about Steve Fielding’s election since it was achieved from major party preferences, it can only be said that he doesn’t know very far. Nobody who was familiar with Green’s position would consider this likely to be the case. And sure enough, if you Google “‘Antony Green’ ‘Steve Fielding'”, the very first thing that appears is a link that gives lie to the assertion.

The election of Steve Fielding did indeed rouse controversy in the group voting ticket system, because it was achieved through a Labor preference arrangement that would have come as a surprise to most of the party’s supporters. Despite Fitzgerald’s conviction to the contrary, dissatisfaction with the system does not begin and end with micro-party preference harvesting, but arises from the fact that voters are very often oblivious to the true effect of their decisions.

As for Brian Harradine, Fitzgerald is, as they say, entitled to his own opinions but not his own facts. Saying Harradine “held federal governments to ransom” falls safely in the former category, but the purported fact that he did so from a 7% primary vote is no less Fitzgerald’s own. In fact, Harradine was first elected with 12.8% of the vote at the 1975 double dissolution, then re-elected with 21.3% in 1980, 13.2% in 1987, 10.4% in 1993 and 7.9% in 1998. In other words, preferences had no bearing on Harradine’s election on the first three occasions, as he achieved a quota in his own right.

When Liberal preferences did help Harradine to a quota at his last two elections, this was only what Liberal voters could reasonably have expected, given the alternative was them going to the Greens. It is for this reason, and not because his preferences came from a major party, that Harradine’s election was not a source of controversy. There is simply no parallel with the results that have lately exercised critics of group voting tickets, such as Ricky Muir’s election in Victoria from 0.5%.

The overarching problem with Fitzgerald’s argument is that he neglects the very important distinction between a system such as that of the New South Wales Legislative Council, in which micro-party members are fairly elected due to what political scientists call the “high magnitude” of its proportional representation, and a group voting ticket system that is everything its critics say it is: “gamed” and “rorted”.

The key distinction here is that New South Wales is not divided into smaller multi-member regions, as the Senate is with its state-based model, and Victoria’s upper house is with its eight five-member regions (which likewise applies to Western Australia’s six-by-six model). At each state election, candidates compete for half the chamber’s 42 seats on a statewide basis. This produces an extremely low quota for election of 4.5%, which becomes a good deal lower in practice because of the very reform advocated by Antony Green – namely, votes do not pass on as preferences beyond the point where the voter expresses a wish for them to do so. Consequently, a quite large share of the vote disappears from the count by the time the final seats are allocated, and these seats are accordingly won from very modest shares of the vote.

So it is that even without group ticket voting, seats for Shooters & Fishers and the Christian Democratic Party, who poll around 3%, are reliable fixtures of New South Wales upper house elections. That these parties in particular should succeed in winning the micro-party seats comes down to a factor to which Fitzgerald appears to (to coin a phrase) be blithely indifferent, namely the number of voters who actually wanted them to win. The system in Victoria proved to be unfair not to the big cartel, but to Palmer United and Shooters & Fishers, who respectively polled 2.7% and 2.3% in Western Victoria, but lost out to Vote 1 Local Jobs with 1.3%.

Having said all that, Fitzgerald does touch upon some genuine issues. Certainly it is true that simply abolishing above-the-line group ticket will raise the “barriers to entry” for new parties, to use a phrase that political science cops from economics in discussing such matters, unless the quota for election is also lowered. To the extent that Antony Green might have advocated one and not the other – and this I’m not sure about – his calls for abolition of group ticket voting can indeed be criticised on such grounds.

In the case of Victoria, the obvious way to resolve the issue would be to abolish the regions and go to a statewide model, such as is used in both New South Wales and South Australia. The same could be said with even greater force for Western Australia, where it would have the further advantage of removing the blight of rural malapportionment. But in the Senate, we run into the brick wall of the Constitution and its imposition of a state-based, low-magnitude model.

Given the scale of the problem uncovered by the 2013 election result, I’m not of a mind to consider this factor decisive. Existing barriers to entry did not stop Brian Harradine and Nick Xenophon winning election to the Senate. The example of Xenophon suggests that aspiring independent Senators could first establish themselves in the state parliament, at least in those states where the barriers to entry were sufficiently modest.

If under-representation of small parties and independents is a problem, having such members elected in the almost entirely arbitrary fashion characteristic of the current system hardly seems the right way of dealing with it. It would make a lot more sense, and be more in line with community expectations, to lower the quota by abolishing staggered terms, so that every Senate election was for 12 members per state, with a relatively modest quota for election of 7.7%.

This would of course require some fairly substantial constitutional surgery, including a return to the question of simultaneous House and Senate elections that was defeated at referenda on three separate occasions in 1974, 1977 and 1984. Things could very well be different if the proposal had bipartisan support, a hurdle it failed to clear on early attempts because an obstructionist Coalition opposed it when the Whitlam government put it forward in 1974, only to advance it themselves when in government three years later. With the Coalition and Labor lining up behind it federally on that occasion, the national yes vote was 62.2%, but the referendum was defeated because it failed to pass by narrow margins in Queensland and Western Australia (no doubt thanks to opposition from Joh Bjelke-Petersen and Charles Court), and a much larger one in Tasmania.

Tangentially, another interesting constitutional question which emerges relates to the double dissolution process, since the outward distinction between a double dissolution and a regular election would cease to exist. If Senate obstruction is to be deemed a problem, perhaps the answer to it is to routinely have the joint sitting mechanism available to re-elected governments in passing the legislation that was blocked in the preceding term.

The Joint Standing Committee on Electoral Matters has today released an interim report on its inquiry into last year’s election, which very pleasingly establishes that the Coalition, Labor and the Greens have agreed to pursue sensible reform to the Senate electoral system. The recommendation is to follow the New South Wales example in having optional preferential voting above-the-line, meaning voters can number as many boxes as they choose and their preferences will exhaust at the point where the numbering of party boxes ends. Those who vote below-the-line will be required to number as many boxes as there are vacancies, meaning six at normal half-Senate elections, twelve at double dissolutions and two at Senate elections for the territories. Crucially, this means an end to group ticket votes, whereby above-the-line voters have a full suite of preferences allocated for them by the party of their choice. As well as closing the door on preference harvesting such as has enabled the election of candidates from as little as 0.5% of the vote, this will discourage the proliferation of micro-parties and the consequent swelling of ballot papers, greatly reducing the very considerable number of voters whose vote does not express their true intention. A further recommendation to make life harder for micro-parties is a requirement that they have 1500 members to register as a party rather than the existing 500, which if anything was less onerous than the equivalent rules at state level and allowed those who cleared the hurdle to field candidates in every state, regardless of how little presence they had there.

The electoral implications of this with respect to the last election have been mapped out by Antony Green, who calculates that “the Coalition would hold 35 seats not 33, Labor 27 not 25, the Greens 9 not 10, and others 5 not 8”. However, this assumes no change in the first preference voting results, when the new system with its less cluttered ballot papers would assuredly have limited such phenomena as the Liberal Democratic Party vote in New South Wales approaching 10% on the back of confusion among Liberal supporters, and voters opting for a micro-party after giving up on locating their true party of choice out of as many as 44 options listed. While it is clear that the new system will make life harder for very small parties, the exhaustion of a large share of the vote due to the optional preferential aspect of the system will mean that the winner of the last seat will usually be elected on well under a full quota, leaving a door open for smaller concerns with a genuinely substantial basis of support. By Antony’s reckoning, “a minor party would probably need about 5% of the vote to have any chance of winning a seat”.

UPDATE: The submissions are now available here, and it turns out the proposal excoriated below is only that of the Liberal Party, and not a major party unity ticket. The Labor Party proposes New South Wales-style optional preferential voting above the line. It is also only the Liberal submission that advocates photo identification when voting.

Dennis Shanahan of The Australian reports that the Liberal Party and the ALP will today make submissions to the Joint Standing Committee on Electoral Reform including, among other things, their recommendations for Senate electoral reform. Interestingly, it appears that both will advocate excluding parties from having their preferences distributed unless they clear a threshold of 1.4% of the primary vote. A common feature of proportional representation systems is that parties must exceed a certain vote threshold in order to win seats, but approaching the issue at the preferences end of the equation is something Sir Humphrey Appleby might have described as “novel”, having never been proposed by any disinterested authority that I am aware of.

While entirely impossible to justify, the appeal of such a proposal to the major parties is simple to discern. It is clear that the election result has led to unstoppable momentum for reform among those with a particular interest in such matters, even if that’s less obviously true of the public at large (an Essential Research poll conducted immediately after the election found 38% thought micro-parties in the Senate good for democracy, with only 25% opting for bad). However, if public submissions to the inquiry are anything to go by, by far the most favoured options involve the abolition of group voting tickets, which in the current manner of their operation are self-evidently an offence against democracy. With the objective of putting electoral outcomes back in the hands of the conscious decisions of voters, the obvious routes to reform are to follow the New South Wales example and distribute preferences only as far as voters purposefully allocate them, or require that voters number every box above the line.

By contrast, a preference threshold alternative could address the controversy directly at hand, namely the election of Senators from as little as 0.5% of the vote in the case of Ricky Muir, while maintaining the major parties’ power to corral preferences to the disadvantage of parties that might in fact clear a conventional election threshold, which are popularly set at around 5% (as is the case in Germany and New Zealand). The instructive case study here is the 1998 election, when the major parties all but froze out One Nation despite their success in polling over a million votes nationally (9.0% of the total). That left them with only one seat out of a parliament of 224, having narrowly achieved a Senate quota in their own right in Queensland. In economic contexts, such behaviour might be described as oligopolistic collusion, whereby established operators act in concert to deny new competitors access to the market.

In fairness, it might yet prove that the parties’ determination to seek an alternative to an election threshold has been inspired by Section 7 of the Constitution and its requirement that the Senate “shall be composed of senators for each State, directly chosen by the people of the State”. In rejecting a challenge to above-the-line Senate voting when it was introduced in 1984, the then Chief Justice of the High Court, Harry Gibbs, allowed that the section required that voters “must vote for the individual candidates whom they wish to choose as senators”, but did not accept that above-the-line voting amounted to anything other than a simplified means for doing so. The principle that the Senate electoral system must be based on choosing candidates rather than parties was nonetheless confirmed, and more than one authority has suggested that this would be violated by a provision that denied election to individual candidates on the basis of their party’s aggregate vote (note that, absent other reforms, candidate-based thresholds would exclude all major party candidates except those at the top of the ticket).

However, it’s very far from clear to me that a preference threshold would get around this, or that there might not be an alternative basis for challenging a system that denied some voters the right to have their preferences considered, while maintaining the privilege for those with the good taste to vote for the parties who are proposing the idea. It’s one thing to penalise a poorly performing party for failing to clear a threshold whose purpose is to ensure that members of parliament do in fact represent a reasonably significant constituency, but a preference threshold seems to place the penalty on those who vote for them. For the sake of clarity of expression, I wrote in the opening paragraph of this post of parties having “their” preferences distributed, which certainly encapsulates the depressing reality of how the system works at present. However, the foundation of any democratic system is that the vote belongs to the voter, and not to the particular party to whom they happen to give their first preference.

Happily, it does appear from Dennis Shanahan’s report that the submissions will grant some credence to the more honest alternative of optional preferential voting above the line. Other proposals said to be “pursued” or “examined” are a national electoral roll for both federal and state elections (difficult to argue with), photo identification when voting (more in the Liberals’ interest than Labor’s, so the precise wording here will be interesting to see), a tightening of the late-campaign advertising blackout, and the no-brainer of prohibiting people from serving as registered officers for more than one party.