We Australians seem to like duopolies or quasi-monopolies. From groceries (Woolworths/Coles), cars (Holden/Falcon), communications (Telstra), print media (News Ltd/Fairfax), arguably banks (cartel of top four) to political parties. As these examples show, such arrangements do not always deliver what is best for the nation or most people.
Now our cozy political duopoly has run into trouble. Thousands of words have been written on the uninspiring election campaign so it will not be revisited here, except to say that it reflected the near impossibility of crafting meaningful policy to appeal across the multicultural, geographically diverse complex organism that modern Australia is today. In reality only a multiplicity of parties can do that.
There has been much wringing of hands about the need for stability but this reaction was either confected or perhaps reflected an understandable nervousness about anything that might disturb the status quo that has, by definition, worked to the advantage of those established interests. There have been a few voices urging everyone there is no need to panic based on some obvious evidence. We have all read many comments on this but bear with me while I attempt to summarize the important points leading to the main contention.
First, the most obvious counter-example is of most European countries which have long had governments formed from multiple and changing party affiliations. No one can argue credibly that Germany shows any ill effects of such a system, being the third or fourth largest economy in the world and the world’s largest exporting nation. The Greens have participated in ruling governments from the late 90s, first in centre-left (so-called red-green) coalitions with the Social Democrats at the federal level and more recently in three-party centre-right coalitions with the conservative Christian Democrats at state level. The respected German foreign minister and deputy PM, Joschka Fischer was a Green.
Second, minority state governments including three in the recent past (Greiner-NSW; Bracks-Vic; Rann-SA) and two today (Barnett-WA and Labor-Greens, Tas) point to the fact that such governments can be stable despite initial expectations. The Tasmanian experiment is only four months old but Labor Premier David Bartlet was extolling his minority government last week and on Insight
Tuesday night, even amidst their first disagreement over betting regulation.
All five key players (two Liberal, three Labor) were in exceptional agreement that, even if uncomfortable at times, these governments were successful.
Third, of the four Anglo democracies of most interest to Australians (USA, UK, Australia, Canada) there has been an effective crisis in governance in the three (USA, UK, Australia) who have inflexible two-party systems. Canada, one of the most successful countries in the world currently has a ministry of right (Conservative) and centre-right (Liberal). In the USA the Tea Partiers, representing less than 1% of Republicans, drive the public agenda of both sides of politics to extremes and into dysfunction. The United Kingdom resolved their hung parliament earlier this year with an unexpected coalition between conservative and centre-left parties, at almost 6 months, shows all the signs that it might be a successful government.
This brief history shows that multi-party politics exists in many of the world’s most successful countries while two-party politics has run into trouble in all countries encumbered with it. Our hung parliament will be resolved shortly and is dependent on four independents whose votes collectively represent 1.23%, but feasibly just one of them might end up being decisive. This means a sole MP with a low of 0.13% (13,681 primary votes, Andrew Wilkie, Denison) to a high of 0.46% (50,366 primary votes, Tony Windsor, New England) of total valid votes cast could determine our government for the next three years.
Although they come from regional Australia -- hardly representative of most Australians -- we seem to be lucky with all four independents acting responsibly, calmly and rationally in the national interest-- imagine if it was someone like Barnaby Joyce, Wilson Tuckey, Bill Heffernan or Steve Fielding! This time we have dodged a bullet.
Nevertheless it can hardly be considered a triumph of our electoral system.
But it is worse. As of Monday the Coalition has moved marginally ahead in two-party preferred terms -- by 1,909 votes or 0.017% of the total. At the time of writing the Coalition are attempting to build a moral case around this miniscule difference. Technically the only thing that is functionally and constitutionally important is that a prospective government must command a majority of seats (76) in the lower house. Morally one would hope that the 1.26 million primary votes of the Greens should matter but their 11.4% share of the vote across the 150 seats in which they ran candidates, won a single seat (Adam Bandt, Melbourne). By contrast, the Nationals, with 3.86% of the vote got seven seats, and the Queensland LNP, with 1.02 million votes (8.95%) achieved 21 seats. An average per seat won of 1.26 million votes versus 48,857 votes. That is a discrepancy of 26 fold.
The main parties and other beneficiaries of this grotesquely skewed result will shrug their shoulders and say that is just the way the cookie crumbles. But it is manifestly unfair and beyond perverse. It is broke and it self-evidently needs fixing.
The last decade of government from both Liberals and Labor has shown paralysis on some critical issues from climate, land management, energy, water, public transport and urban planning. There are plenty of solutions but the impediment to their implementation is political and the apparent loss of political validity of the major parties. It seems clear we need to resolve that by breaking up the monolithic parties into smaller units that can honestly represent the diversity of options available. It is interesting that the resolution to the election and many of the divisive policies lies outside the main parties, by parliamentarians who years ago came to the conclusion that the solutions did not lie within those parties. But a proliferation of independents (which may well be on the cards) cannot by itself lead to a long term resolution and it seems the only way to achieve this is via electoral reform.
Electoral reform is extremely difficult to introduce but not impossible. Entrenched interests of major parties, conservatism and inertia all work against it. Hare-Clark is a proportional representational (PR) voting system—with multi-member electorates—that combines features of both PR and a preferential system (otherwise known as Single Transferable Vote, STV). The original Hare-Clark PR-STV system was introduced very early in Tasmania in 1896 but the popularity of versions of PR-STV in the modern world can be judged by its adoption in totally new constituencies such as Scotland’s and Northern Ireland’s regional governments and the Australian Capital Territory.
The UK election in March this year is a clear model for what can be wrong about electoral systems. The two main parties got 36.1% and 29% (total 65.1%) of the vote but in their first-past-the-post system it won them 87% of the seats. Neither party obtained a majority to be able to form a government. The Liberal-Democrats won 23% of the vote yet obtained only 8.8% of the seats. It has become generally accepted that there needs to be electoral reform but the graphic shows that if the UK adopts our version of preferential (Alternative) voting this extreme inequality in distribution of seats will only modestly be corrected. Any Australian could have told them! On the other hand if they adopted PR-STV (see graph) the political situation is transformed.
The Australian data in the graph shows a similar effect if this election’s data are modeled* on a Hare-Clark system. As noted* this is a very broadbrush view. The increases or decreases in Liberal and LNQ seat tally presumably reflects current concentration of those votes in specific locations which is not accurately accounted for by the national or state averaging performed here. As with today’s outcome no single party has a majority.
The defining mechanism of Hare-Clark is the quota (or threshold) of votes required for a candidate to be elected: if 100 valid votes are cast then the quota is: 100 divided by “number of seats plus 1”, plus 1. Thus in a 5 seat system the quota is 16.7+1=17.7 ie. 18 votes; or more generally “16.7% of votes plus one vote”—for example if 1000 valid votes are cast the quota is 168 votes (167+1).
A strong feature of STV is to minimize “wasted votes”. It does this by transferring surplus votes from those candidates who have reached the quota; that candidate’s surplus votes are transferred in proportion to the overall second choice preferences of that winning candidate. Thus if in an election for 5 seats with 100 valid votes cast, requiring a quota of 16.7% (ie. 18 votes) if one candidate receives 50 votes then the 32 surplus votes are transferred to the second-preference candidates according to the overall preference of that candidate’s 50 voters.
The process is repeated until all seats are filled or, if unfilled, then the lowest-polling candidate is eliminated with their votes being redistributed to their second preferences.
While this may be complicated to explain, and is criticized as impractical for that reason, in the voting booth it is little different to our current Senate PR voting system. It is admirably simple and clear: the voter numbers 1 to 5 their top choice of candidates. In STV there is no party list and no “above the line” voting. Strong independent candidates, such as the four currently determining our future, have nothing to fear from such a system.
The five member system is the preferred choice because it is a balance between reasonable representation of “significant” minorities (requiring, after allocation of preferences, 16.7% of all votes) and simplicity and practicality. It has the added benefit of always voting for individuals but the larger the number of seats per electorate, the larger the electorate needs to be and thus a diminution of “local” representation.
So, is it very likely that we might use this exceptional opportunity to evolve our electoral system to be both fair and more likely to cope with our increasingly complex society in an increasingly integrated complex, changing world? No, of course not.
With responsible independents in the lower house we have a real chance at improved quality of governance. In the coming months and years the Greens with their balance of power in the Senate will have their chance to express their political power. The public will probably not give them a second chance if they refuse to be pragmatic and flexible and deign to be obstructionist.
Most of all we might hope that the media and commentariat might accept that our creaky old electoral system is broke and is at least partly responsible for our political system being so dysfunctional. A multi-party system more accurately allows honest expression of policy differences rather than the grey soup we currently have from attempts to not upset anyone.
In the past Australia has been a world leader in women’s right to vote, protection of workers rights, the first Labor party and the first national preferential voting system. Once again we could show the world how to evolve a fairer and responsive democracy.
*Footnote on Methodology: Calculating outcomes with Hare-Clark models is simple with actual multimember polling data but very problematic to impose on existing polling data, for three main reasons. 1) Voter behaviour: this will change considerably due to both local candidates carrying more weight and perception of less “waste” of their vote for minor parties and independents. Tasmania proves both points and led to a Labor-Green government. 2) Preference allocation: with multi-member electorates voters are faced with very different choices than in our two party system; however giving the main parties the benefit of the doubt we assume the main parties will preference either their second candidate 75%, or in the case of the Greens and Nationals (if the Nats can still be called a major party) the second preference will go to Labor and Liberal respectively. 3) A third factor is geography which would require complex highly-locale specific data and assumptions about make-up of the much bigger electorates; it is not attempted here.
As a simplistic broad-brush look, the formal vote totals were divided by the calculated quota. Due to the LNP in Queensland that state was modeled separately to the rest of the country. Data for the 2010 election was downloaded from the AEC on Saturday 28 August and partly updated Monday 30 August.
Dr Michael R. James is a research scientist and writer. He is not a member of any political party.