Electoral analysis has shown that preference voting would favour Labour and the Liberal Democrats, whose voters already employ tactical voting to defeat Conservative candidates by voting for whichever candidate is in a stronger position. Rather than producing a proportional result, it would have resulted in an even larger Labour majority in 1997 when they did not come close to winning a majority.

Unsurprisingly, the Conservatives have come out strongly against the proposal and continue to support the first-past-the-post system, while the Liberal Democrats have argued that the proposal does not go far enough.

The Conservatives have a solid lead in polls for the election, which is expected in May or June, but the electoral system means that a large lead is needed for the party to win a majority, suggesting a strong possibility of a hung parliament with the Liberal Democrats and parties from Scotland, Wales and Northern Ireland sharing the balance of power. This means that it is plausible that negotiation following the election may revolve around kick-starting the electoral reform process, with Labour now committed to a first step and the Liberal Democrats insisting on proportional representation as a key priority.

The legislation will be passed before the election, which would mean the referendum could go ahead regardless of who won, although it is conceivable that a Conservative government could call the referendum to a halt or a deal with the Liberal Democrats could see the scope expanded.

They also carry an article from NSW Opposition Leader Barry O’Farrell pushing the proposal and a petition for a referendum at the next election to “reclaim your vote”, something completely bizarre and pointless, since such a referendum would not hasten the end of the current Labor government and O’Farrell has promised a referendum in his first term.

There’s nothing wrong with the idea of including a recall provision in our constitution, but it is a hugely costly and ineffective way of achieving greater accountability from our state governments while the Herald and O’Farrell ignore far easier and more effective ways of holding bad governments to account.

Surely the push for recall elections comes from a sense that state governments are too distant from the people and that we need to restore accountability in government. If you want to achieve that, recall elections only create an illusion of accountability while keeping the power of government locked up in one-party government.

Recall elections are extremely costly and ineffective. As O’Farrell admits in his piece, only two governors in US states have been recalled in the last century. While recall elections can very rarely provide accountability in the case of incredibly unpopular and incompetent governments, most run-of-the-mill governments completely avoid any fear of the public being able to bring them into account. Don’t even get me started on the idea that we would model our system of government on California, one of the most dysfunctional political systems in the developed English-speaking world.

The problems of bad government usually can’t be solved by holding an election. It’s widely agreed that the Labor government was already acting appallingly on many public policy issues before the 2003 and 2007 elections, yet they were reelected in landslides at both elections.

Most of the time the public are disengaged and won’t act politically to deal with day-to-day issues of bad government. The recall process means ‘business as usual’ until the public are willing to rise up en masse and throw out a government.

If you want to have real accountability, there are a few ways you can do that. Firstly, we should surely be talking about reducing parliamentary terms from four years to three, as they were until the 1980s. Barry O’Farrell engages in the idiotic meme blaming NSW’s current woes on the ‘fixed term’ system:

They argued that fixed terms would end community frustration with the calling of early elections and allow government the certainty needed to get on with the job of governing.

But at that time no one foresaw how a NSW Labor machine could, and would, pervert the fixed term system to protect maladministration, political inertia and incompetence.

Of course, if we didn’t have fixed terms in New South Wales, nothing would have changed. We still would have an ageing state government because, as O’Farrell acknowledges later in his piece, they know they would lose and they would not call an election. Isn’t it interesting that O’Farrell doesn’t promote three-year terms? Indeed, his proposal of keeping four-year terms but abolishing fixed terms would increase the power of the Premier when he takes office.

The other solution that would genuinely increase government accountability and would see bad governments like our current government terminated mid-term would be to introduce proportional representation in the Legislative Assembly. By doing so, one-party majorities would only be won in rare landslides, and most Parliaments would not be controlled by a single party. A major party would need to convince minor parties and independents to give their support and possibly govern in coalition in order to rule.

Where a government acts terribly and the cries for an early election grow loudly, or where a party replaces their leader with an unknown or unpopular figure, it is possible for their support parties to withdraw support and either form a different government or force an early election, and they can do it much easier than a mass petition of hundreds of thousands of voters.

The current Labor government would not have won a majority under proportional representation. The balance between the two major parties would be better, and Labor would have had to rely on the Greens and independents in order to govern. I would expect that, if the government was still as bad as it is now, their support parties would have pulled the plug early and forced an election. At the very least, they would prevent the party governing as an elected dictatorship. The best thing about PR-fuelled hung parliaments is that you have democratic accountability of all governments, not just those on the brink of death. It is far more effective and useful than a recall petition.

Indeed, the Canadian province of British Columbia introduced the recall election in 1995, yet this did not help introduce any more accountability into the political system, and in recent years the province has considered the question of proportional representation as a solution to the problem of massive government majorities blocking accountability.

If you want to reduce the power of governments to continue to govern without the support of the public, you don’t need recall elections. You need proportional representation in the lower house and fixed three-year terms.

In addition, submissions will close on November 27 for those interested in commenting on the Green Paper. I plan to put in a submission, and hopefully I’ll be able to post some ideas later this week, but I thought commenters might have their own suggestions about what to put in submissions to the Green Paper.

In addition there are candidates for the Liberal Party, Greens, Australian Sex Party, One Nation, Liberal Democrats, Democratic Labor Party, Environmentalists for Nuclear Energy, Climate Change Coalition and five independents.

The implications of this are dire for the conduct of the by-election. Compulsory preferential voting means that all voters will need to express a preference for every single candidate, including expressing a candidate between all nine Christian Democrats and many independents and political parties that few Bradfield voters would have heard of.

Voters who failed to express 22 preferences or make mistakes can have their votes declared informal, even if they express a clear preference. These voters may have voted for the Liberals or Greens, and their votes would never be passed on to another candidate, yet these votes would be thrown out for an irrelevant error.

I believe there is a rule which allows 10%  of preferences to make mistakes, which would allow a voter to make two mistakes in the Bradfield by-election. (Update: turns out you are only allowed to make one mistake. The 10% rule only applies to the Senate). Even still, you can expect a huge informal vote.

At the 2005 Werriwa by-election, there were sixteen candidates, and this election produced an informal vote of over 13%. The informal vote was higher than all candidates other than the winning ALP candidate, Chris Hayes, who polled 55%. No other candidate polled over 10%.

It’s not unreasonable to think that the informal vote could break 20% in the coming by-election. In addition, it will be extremely difficult for parties to produce how-to-votes that allow people to vote easily. Take the Greens for example. Do you use principle in preferencing, by putting One Nation last and trying to distinguish between all of the right-wing minor parties and independents, or do you simply produce a donkey vote HTV to ensure your voters have their votes counted?

The solution is very simple: some sort of optional preferential voting. I personally think that there is no justfication for tossing out a vote for expressing an invalid lower preference. If we cannot achieve full optional preferential voting, there should be at least a cap on the number of preferences that need to be expressed.

If it was said you had to number at least five preferences it would encourage parties to preference and would ensure preferences flow when there are a small number of candidates, whilst avoiding high informal votes in situations like Bradfield and where large numbers of candidates run in a general election (fourteen candidates ran in Greenway in 2007).

In addition, of course, another reform suggested by Antony Green should be adopted. This would be that, while parties are permitted to nominate multiple candidates, any additional candidates would be required to go out and get signatures from voters, just as independents are required to do.

I haven’t really delved here into the question of whether it makes sense for the Christian Democratic Party to run nine candidates. While it could result in a slightly higher total CDP vote due to each candidate getting the support of their own community, it’s undoubtedly going to hurt them.

They would have needed to spend $5500 $4500 on nomination fees, and it seems impossible that any candidates could individually poll the 4% needed to qualify for public funding. In such a large field of right-wing candidates, all of whom will be competing with an official Liberal candidate, you can’t see any of them doing particularly well.

British Columbian politics is particularly unusual in Canada. The two major parties are the Liberal Party and the New Democratic Party, but they fill very different roles in the provincial political environment. The Liberal Party is the centre-right party, and has no affiliation to the federal party. Many BC Liberals are aligned with the Conservative Party of Canada. The NDP fills a more centrist role than in federal politics.

The Liberals, under Premier Gordon Campbell, have governed BC since the 2001 election, when the ruling NDP lost all but two of their seats, with the Liberals winning 77 of 79 seats. This was the culmination of half a century of volatile provincial politics.

BC’s peculiar political system dates back to the 1940s, when the major Liberal and Conservative parties formed a coalition to prevent the left-wing Cooperative Commonwealth Federation, the predecessor to the NDP, from gaining power. The 1952 election saw the coalition split and saw the fringe Social Credit League form a minority government. The CCF was in opposition with the Liberals and the Conservatives reduced to the crossbenches. Social Credit was founded to pursue the policies of social credit parties, who remained a fringe movement in most of the English-speaking world.

The party quickly broadened into a centre-right conservative party, with the CCF, then the NDP, becoming the main centre-left party, and the Liberals and Conservatives almost entirely disappearing from BC politics. This state of affairs continuing for the next thirty-nine years, with the exception of three years of NDP government in the 1970s. In 1991, the Social Credit party was decimated, not only losing the election to the NDP, but being pushed into third place behind the resurgent Liberals. The Socreds disappeared at the 1996 election, and the Liberals defeated the NDP in the 2001 election. 2005 saw a resurgent NDP gain ground but fail to unseat Campbell. The province has one of the strongest provincial Green parties, with the Greens peaking at 12% at the 2001 election, although they went back slightly in 2005.

The huge margin of victory in 2001 was followed by the Liberals carrying out a promise to create a Citizens’ Assembly on Electoral Reform. The body was made up of randomly selected citizens selected from all provincial ridings. After a period of examination of the various options available, the Assembly approved a variant of the Single Transferable Vote system used in Ireland and Tasmania (known in Australia as Hare-Clark). The proposal was taken to the 2005 provincial election as a referendum, and managed to gain 57% support, as well as majority support in all but two ridings. However the legislation required the referendum to win 60% support, and thus it was defeated. Gordon Campbell’s government has pushed ahead with their plans to introduce BC-STV, and will introduce another referendum alongside the election in 2009, again with a 60% threshold imposed.

British Columbia will also be voting using new electoral boundaries in 2009, and as part of the boundary review process the Commission proposed a set of boundaries to be used in an STV election. These boundaries combine single-member ridings together. For example, five ridings in eastern Vancouver will be combined to make Vancouver East riding, which will elect 5 MLAs as a group. I have converted these files to Google Earth and uploaded them to my maps page. However, I haven’t been able to find any calculations of the notional margins on new boundaries. I would appreciate it if anyone has information on that front.

The prospect of British Columbia adopting the Hare-Clark system is exciting for anyone interested in electoral reform. A successful implementation of BC-STV would not only influence fellow Canadian provinces, and the Canadian federal system, which also suffer from similar problems, it could also see an impact on the neighbouring US west coast, where relatively progressive regimes in California, Oregon and Washington have been the most eager to experiment with new electoral systems. The province of Ontario has already adopted BC’s Citizens’ Assembly model, with an Assembly proposing an MMP system similar to New Zealand, which was soundly defeated at a referendum in 2007. While this failed, victory in BC in 2009 would be a big step towards the end of First Past the Post elections.

It appears that the plan is to use the system in one of a number of traditional Nationals seats held by a rural independent, such as Dubbo, Port Macquarie or Tamworth. It appears a smart strategy to blunt the impact of rural independents, and in certain cases would prevent cases of popular candidates being defeated by party machinists. As van Onselen points out in the Australian article, federal member for New England Tony Windsor would have likely won the Nationals primary for Tamworth back at the beginning of his career, and would have remained within the party.

If such a model spread through politics, it would have a fascinating impact. MPs would be much less beholden to their parties and we would likely see a decline in party discipline. It could also have a serious impact on government ministers. Yet it seems unclear how a primary system can effectively work in a political system which isn’t strictly divided into two parties, and it is completely incompatible with any system of multi-member election system.

It would seem to be a step in the right direction, but it would make more sense to give more powers to “one vote one value” elections within the party, which would be a strong incentive to encourage more voters to join political parties, while avoiding the obviously silly concept of voters from the opposite end of the spectrum having a say over a party’s candidates. The Nationals have a very large membership base, and it would seem to be just as effective to give the power of preselection to a vote of all members living in the electorate. It would seem bizarre that Labor and Greens members in, say, Dubbo, let alone supporters of the sitting independent, could have a say over who the Nationals stand.

While an open primary system may not become the universal system of preselecting candidates, it is a good gimmick and can be useful for the Nationals in regaining momentum in country areas which have become disengaged from the party. The rise of maverick Nationals who are more concerned with the party’s independence than its coalition relationship, such as Barnaby Joyce and Brendan Grylls, would be encouraged by the rise of open primary preselections.

Update: That dangerous lunatic Tim Andrews has some unkind words to say about this post over at his blog. Check it out.