Electoral systems Archive

Abolishing the Electoral College?

Via FiveThirtyEight, there’s an interesting article in the Wall Street Journal outlining efforts to effectively abolish the Electoral College in the US without changing the constitution.

This approach, called the National Popular Vote Interstate Compact, would see states pass legislation committing them to a binding agreement that they would cast all of their electoral votes for the winner of the national popular vote. The compact won’t come into effect until states which hold a majority of electoral votes have signed on. The WSJ article continues:

The debate hits full stride now in Colorado, a state that political analysts say presents a key test for the National Popular Vote project. So far, the states most receptive to doing away with the Electoral College have all been solidly Democratic — not the swing states that have been high-profile players in presidential elections.

But Colorado last year joined a small cluster of newly minted swing states that drew a disproportionate share of candidate visits and campaign spending. It will now help answer the question of whether swing states will take the leap.

As Nate Silver points out, only 50 EVs so far have been committed to the compact, and all the states who have passed the legislation or close to passing it have been very Democratic and very safe states when it comes to presidential elections.

Nate Silver further argues that there is a high hurdle to be jumped before any electoral college reform can gather enough steam to be passed:

What would it take for there to be a real chance of abolishing (or end-arounding, as the Compact seeks to do) the Electoral College? I think it would take two elections in relatively rapid succession in which there’s a popular:electoral split, particularly if these two elections are won by candidates of opposite parties. The memories of 2000 should linger for a few more cycles, and so if there’s another such occurrence before, say, 2020 or 2024, things could get very interesting.

Apart from the issues in getting it passed, the biggest barrier to electing the US President by popular vote is the current US system of election administration. Unlike the Australian Electoral Commission, elections are conducted on a very local level in the US, and votes are never counted across state boundaries. Indeed, there is no official record of the popular vote, it is simply tallied by media organisations based on state tallies of the popular vote.

In the case of an extremely close national vote, the issues that popped up in Florida in 2000 would take place on a massively expanded scale. On the other hand, a directly-elected presidency could be just the thing the USA needs to finally bring election administration completely under federal control.

Psephoblogs around the world

I love reading all the psephological blogs in Australia (primarily Crikey’s Poll Bludger and Possum), as well as the bits of psephological content on a bunch of broader political blogs. However, it’s never enough, and I like to read psephological blogs from other English-speaking countries.

I’m always looking for more, and would like to make more links and gain insight into how elections are covered in other countries, so I’m gonna list some of my favourite foreign psephoblogs. What ones do you read? Please go to comments and post your favourites for everyone else to read. And if you know any English-language psephoblogs covering elections in non-English-speaking countries, I would love to see it.

  • Curiablog – A relatively quiet New Zealand blog that posts regularly covering New Zealand opinion polls, and provided a polling average during last year’s NZ election.
  • FiveThirtyEight.com – The star of the 2008 US Presidential election, FiveThirtyEight was only established in March 2008, well into the election campaign, but Nate Silver’s unique statistical analysis, adopted from baseball fandom, managed to make him a major psephological pundit by the time of the general election in November, and was incredibly accurate in projecting the result.
  • Irish Election – A blog I’ve just discovered, it seems to be the Poll Bludger of Ireland, and would be worth focusing on, should Ireland’s government collapse and trigger an early election in the months to come.
  • UK Polling Report – UK Polling Report provides analysis of UK opinion polls and has produced impressive analysis of the seats that matter for the next UK general election. It includes a calculator that predicts seat numbers based on the percentage of the vote going to each of the three major parties.

Israeli electoral reform?

Israel is one of the only countries in the world where its entire legislature is elected to represent the country as a whole, and it is often blamed for Israel’s unstable and rapidly-changing party system. It is true that Israel’s pure form of proportional representation contributes to instability in Israeli politics, although the dragging on of the Arab-Israeli conflict and Israel’s unique ethnic diversity and geopolitical position in the Middle East suggest that no electoral system would shift towards a stable two-party system any time soon.

In response to this week’s election results, there have been numerous calls for electoral reform in Israel, incuding from Tzipi Livni and Yisrael Beiteinu leader Avigdor Lieberman.

The ABC quotes former Labour MK Professor Shimon Shitreet proposing a possible model:

Professor Shitreet and others are proposing a change they think the smaller parties will accept, where half of the members would be elected by district and half would remain under the proportional system.

This model is used in Japan, South Korea, and most interestingly elections to the Palestinian Legislative Council, the body that became dominated by Hamas at the 2006 election.

Another possible model would be the MMP system used in New Zealand and Germany. Using New Zealand’s system, all parties except the five largest would be wiped out, making it easier to form a government. However, it is likely that at least some Arab parties would win constituencies in the north. The district around Acre would have four districts out of sixty, and has a substantial Arab population. This would allow Arab voters to keep representation in Parliament.

The biggest obstacle is that any electoral reform would almost certainly result in a higher electoral threshold, even if that comes about through indirect means, such as electing MKs through multi-member constituencies. While the five largest parties, as well as the Arab parties collectively, would still win seats under most mixed systems, the smaller parties would likely be wiped out or forced to merge with each other or with a major party. Most scenarios for forming a new government include some of those parties in a new governing coalition, and in spite of their small numbers they would likely veto any potential electoral reform. However, a potential alliance of Likud, Kadima and Yisrael Beiteinu would provide a majority government of large parties that could introduce serious electoral reform.

The Redistricting Game

A weekend treat for all you electoral nuts out there: I’ve just discovered the Redistricting Game. It’s a net-based computer game built by people campaigning for fair electoral districts in the US. Basically you play as the person appointed to draw boundaries in the House of Representatives for the majority in the state legislature. You’re given four congressional districts, and have various tasks: simply accomodate changes in population density, gerrymander the seats to give your party three of the four seats, or gerrymander them to produce two safe Republican seats and two safe Democratic seats. A lot of fun.

How do we elect our leaders? Part two

Yesterday I laid out how leaders are elected in Canada and the UK, where party members play a central role in deciding leaders, as opposed to New Zealand and Australia, where leaders are elected by members of Parliament.

So what are the merits of the alternatives? Whenever you raise the possibility of electing party leaders by the grassroots membership, you’ll be referred to the patchy history of the Australian Democrats electing its leaders through a general election of the membership. It’s true that the conflict between Meg Lees and Natasha Stott Despoja was dragged out through a long election process, but it remains that the party room was deeply split and such a conflict probably would have continued even if the Senators had responsibility for electing leaders.

When you look at how direct election processes have elected leaders like Tony Blair, David Cameron and most recent Canadian political leaders, you would have to think that precedent suggests that a direct election process does not necessarily lead to instability and internal division any more than a process that gives the final say to a parliamentary caucus.

Issues of giving control of leadership to party members in a direct vote has come up a few times in recent years. Following the 2004 election, the Australian Greens officially appointed Senator Bob Brown as Leader after almost a decade a de facto leader. Although there was little doubt that Brown would be elected, the process of electing leader was controversial, and Brown’s favoured model of giving total control to the Party Room was approved, although some called for the decision to be made by the National Council or the membership.

In 2003, the National Presidency of the Australian Labor Party was put to the membership in a direct vote for the first time, a reform initiative of then-leader Simon Crean. With a turnout of just under 20,000 members, former WA Premier Carmen Lawrence, of the Left, was elected as President. The position is elected every three years, with the second- and third-polling candidates serving as Vice-President for the first year after the election. The Presidency is then rotated, so that the three top-polling candidates each serve one year as President and two years as Vice-President. The following election in 2006 also saw a Left candidate win, with Senator John Faulkner elected President for the year 2007, defeating SA Premier Mike Rann amongst others. Those results suggest that, despite the domination of the Right faction, ALP members are substantially more left-wing than their leadership. Which might not be that much of a surprise.

After years of internal fighting, the Victorian Liberal Party has recently implemented new preselection rules which give all members in an electorate a vote in deciding candidates for preselections. This has been seen in the upcoming campaign to succeed Petro Georgiou in Kooyong. Rather than a handful of senior party members, over 1000 party members living in Kooyong will vote in a small-scale primary to decide the Liberal candidate for the 2010 election. While not quite the same as giving members the right to decide the leadership, the principle of giving all members the right to decide their leaders in order to cut through internal wrangling remains the same.

So what are the merits of the alternatives available? The 2008 Democratic presidential primary in the United States clearly demonstrated the merits of a real election within a party. Although many worried that the lengthened Democratic race would favour McCain, it ultimately strengthened Obama as a candidate, built up Democratic membership and machines across the country and invested Democratic voters in the Obama campaign. Despite fears that division in the party would drive Clinton supporters away from the party, the issue was highly over-rated, and division was far more of a problem for McCain than Obama. Of course, no leadership election in the UK or Canada comes close to the proportion of voters participating in US primaries, but the principle is the same. A real election will beat a focus group or opinion poll in proving a candidate’s worth any day.

History shows that Canadian and UK leadership elections inspire parties in a way that nothing else does. When an inspiring figure is a candidate, or a debate over the direction of the party is taking place, a large-scale election can work like nothing else. Parties recruit large numbers of new members who become organised and politicised in a way that is hard to do otherwise. The New Zealand Greens’ 2006 leadership race provided an opportunity for the party to debate the way forward: should they remain committed to supporting a Labour government, or would their agenda be better advanced by being flexible and being able to work with the Nationals? Most importantly, it makes for better leaders. There can be plenty of ulterior motives for MPs or senior officials to support a particular leader despite not being the best person for the job. But no-one can win a large-scale election without having a decent level of campaign ability.

Of course, there are arguments against allowing voters and party members a say in deciding their leaders. A party’s membership base tends to be more extreme than the voter base, and can support candidates with less popular appeal. Successful opposition leaders tend to be elected by expanding their party’s voter base rather than simply relying on their existing support base. But you have to wonder why such a narrow group of Australians are represented in today’s political party memberships. Surely it has something to do with the complete lack of influence in major parties given to the grassroots membership? If you gave members of, say, the ALP, a significant say over who is the leader of the party, it would be a greater motivation to join than anything else you could offer. It would undoubtedly broaden the base of the party. And if there is one group of people less politically diverse than members of political parties, it’s the MPs and officials that run those parties.

The one complaint heard most when people object to letting more people have a say in electing party leaders, it is that a widespread election process airs out private party matters and internal divisions, which of course is true. You can’t expect that issues raised in a process involving thousands or tens of thousands of members to remain private. But if you look at recent contested party leadership races, such as the ongoing battles between Kim Beazley and Simon Crean, Mark Latham and Kevin Rudd respectively, or the ongoing Turnbull-Nelson skirmishes, it’s clear that little in terms of internal party divisions ever remains private, regardless of the method of election. If MPs can leak sensitive party matters without the party collapsing into public disrepute, then it makes sense that no more damage can be done by an open and genuinely democratic process.

It’s a big culture shock in Australian politics to suggest that our country’s political leadership is chosen not in smoke-filled rooms but in openly democratic processes that encourage many more Australians to actively participate in elections where their votes genuinely matter. It could be the best solution to the widespread dissatisfaction with modern Australian democracy that we could find.

I thought this would be a good opportunity to test out WordPress’ polling mechanism, so I thought I’d post one to see what people think. Please post your comments about what I’ve argued. Do you think we need a change in how we elect our leaders?

[polldaddy poll=1195782]

How do we elect our leaders? Part one

After the events of the last few weeks further demonstrated the inability of Stephane Dion to remain as leader of the Liberal Party of Canada, Dion announced last Monday that he would resign as soon as the party chose an interim leader. Dion had announced his resignation following the October election, and a leadership convention was planned for May, with three contenders.

On Monday 8th, Dominic LeBlanc withdrew and supported Michael Ignatieff, and Ignatieff’s main rival, former NDP Ontario Premier Bob Rae, withdrew on Tuesday, giving room for Ignatieff to be elected interim Leader by the Liberal Party executive in consultation with the caucus on Wednesday 10th December. The convention will still be held in May, but is expected to be a coronation, with Ignatieff solidly in place as Leader of the Opposition or even Prime Minister.

The difficulty the Liberal Party had in reconciling its lengthy leadership process with the need to make a quick decisions raises some interesting issues about how political parties elect their leaders.

Canadian parties use various methods of allowing their members to have a say in electing members. The Liberal Party elect leaders at a convention that resembles old-fashioned US presidential conventions, where candidates are gradually knocked out until one gets support of the majority. The last convention in 2006 had about 2600 delegates voting at the convention. The NDP, BQ and the Conservatives all use various processes that give all members a vote. The NDP gives 75% of votes to members of the party, and 25% of votes to members of affiliated organisations, which are mainly labour unions. Over 58,000 votes were cast at the last leadership election in 2003. The Conservatives use a weighted system that gives 100 points to each of 308 ridings. The ridings are distributed proportionally according to how the members living in that riding voted. The BQ appears to use a simple “one vote one value” system, and the last leadership election saw about 48,000 members vote. Most Canadian provincial parties also seem to have shifted towards a “one vote one value” system as well.

UK political parties likewise use various systems that put the ultimate say largely in the hands of grassroots members while giving some say to Members of Parliament. The British Labour Party uses a system which weights votes so that 1/3 of the vote is cast by members of the constituency parties, 1/3 by members of affiliated organisations (mainly labour unions) and 1/3 by members of Parliamentary Labour Party. In the only contested leadership election in 1994, Tony Blair won 57% of the vote, with a majority in all three parts of the electoral college.

The British Conservative Party uses a process whereby candidates face voting by Members of Parliament until there are only two candidates remaining, and then the two proceed to a vote of all grassroots members. In the 2005 election, four candidates nominated. David Davis received the most votes in the first round, but Cameron took the clear lead after the lowest-polling candidate was eliminated. In the members’ vote, almost 200,000 votes were cast, and Cameron won clearly with 67% of the vote.

The Liberal Democrats use a straight one-vote-one-value process, although each candidate must be nominated by at least 10% of the parliamentary party. Members vote with a preference ballot. Following the 2006 resignation of Charles Kennedy, 52,000 members voted, with Menzies Campbell winning 44% of the primary vote, being elected on preferences, beating Chris Huhne with 58% of the preference vote. Another leadership election in 2007 saw Nick Clegg beat Chris Huhne by a slim margin of 511 votes out of 41,000 cast.

In contrast, Australia and New Zealand politics tends towards members of Parliament deciding leaders. The only exceptions I can find are the NZ Green Party and the Australian Democrats, although only two state Greens parties have official leaders (ACT and Tasmania), so Greens MPs sit in Parliaments in NSW, Victoria, Queensland, Western Australia and South Australia without any elected leader.

The New Zealand Green Party has two co-leaders, and they are elected by delegates to the national conference of the party. Their constitution requires that one leader be male and the other female. The party’s original leaders, Rod Donald and Jeanette Fitzsimmons, remained in place for the 1996, 1999, 2002 and 2005 elections, and the only recent election took place in 2006 following the death of Rod Donald shortly after the 2005 election. The election was contested by MP Nandor Tanczos, former MP Mike Ward and party members Russel Norman and David Clendon. Due to the fact that four of the six Greens MPs were female, there was a high chance that the new male co-leader would not be an MP, and in the end Norman defeated Tanczos in a preference ballot. Norman became leader outside Parliament and was elected to Parliament to fill Tanczos’ seat when he resigned in mid-2008, and Norman was re-elected at the 2008 election.

I was planning on going into what we should do in Australia as far as electing our leaders, but this has gotten too long, so:

Tomorrow: what should we do in Australia? What would happen to our politics if grassroots members got to decide who became party leader?

Italy’s overseas seats

ABC Radio National’s Background Briefing program broadcast a fascinating program two weekends ago regarding the interaction of Italian and Australian politics arising from Italy’s expatriate seats.

The 2006 Italian election was the first election since the Italian electoral system was changed to include seats in Parliament dedicated to Italian citizens living abroad. This system divides the entire world into a number of electorates, including the “Oceania” electorate, which elects one Italian MP to the Chamber of Deputies and one Senator, in an area covering Australia, New Zealand, the Pacific Islands, Asia, Africa and Antarctica. Despite the vast expanse of the electorate, most live voters live in Australia, particularly in Melbourne.

The program lays out the fascinating complications added to the political system when elections spill over national boundaries. In particular, the centre-left coalition L’Unione came into conflict with Italian-Australians and particularly Italian-Australian members of the Australian Labor Party through the Italian-Australian Labor Network. Figures such as Victorian state MP Carlo Carli and Mayor of Moreland Joe Caputto. The IALN signed a factional agreement with Italy’s centre-left parties agreeing that Italian-Australian ALP members would support the centre-left candidates in exchange for control over future preselections.

It raises interesting questions about its effectiveness, the ethics of giving citizens living abroad a say over the government of a country they no longer live in, and how it could be used for other countries.

In New Zealand, the NZ Green Party has repeatedly nominated a handful of NZ expatriates living in Australia and the UK, some of which are active in the local Green parties. Maori Party co-leader Pita Sharples recently proposed that, now that one-eighth of Maoris live in Australia, an eighth Maori seat should be established in Australia. How would this inter-relate with Australian politics, particularly in Queensland where a majority of Maori expatriates live? What about the large numbers of New Zealanders living in Australia?

Over one million Australians live overseas. Should electorates be established in the UK, China, Japan, the United States? It might seem unconstitutional at first glance, but then again there is no reference to Territory electorates in the constitution either. The concept of a Member of the House of Representatives representing Australians in China raises the issue of how a foreign democratic election could be conducted in a dictatorship.

What do you think? How do we represent the hundreds of thousands of Australian voters disenfranchised by living overseas? How would Australian politics be influenced by electing MPs representing expatriates?

Reflections on MMP

Malcolm Mackerras has written in Thursday’s Crikey, critiquing New Zealand’s Mixed Member Proportional electoral system (subscription required):

Having studied the referendum results of September 1992 and November 1993, and having studied opinion poll data, it is clear to me that a solid majority of New Zealanders believes that the system of single-member electorates is a fair and reasonable way by which to elect 70 members of their parliament. That being so it would be fair to allow political parties to get the full benefit of their winnings of such seats.

However, MMP does not do that. The Maori Party, United Future and Act are allowed to get the full benefit but, under this contrivance, Labour and National are not. For every extra constituency seat Labour and National wins they are robbed of a party list seat.

He also says:

The ballot paper handed to each elector reads “You have 2 votes”. However, ordinary New Zealanders are smart enough to know what that means. If you are on the Maori roll you really do have two votes. Again if you are on the roll for Epsom, Ohariu or Wigram you really do have two votes. For everyone else, however, it is a two-ticks-one-vote system. Your one real vote is your party vote.

Mackerras’ argument is that the concept of the electoral threshold means that, in those electorates where a minor party is competitive, the race for an electorate becomes a significant contest, while any other electorate becomes pretty much meaningless. Mackerras is also a supporter of PR-STV or Hare-Clark, so he isn’t simply an opponent of PR systems.

There are two main problems with New Zealand’s electoral system. The first is the fact that certain seats still retain significance, but rather than the marginal seat problem in single-member electorates, where most electorates are safe, electorates take on greater importance in New Zealand where they make the difference between a party making it into Parliament or not. At the 2005 election, ACT leader and Epsom candidate Rodney Hide made the argument to the conservative electorate that, by voting for him, the electorate would still see his National opponent elected as a List MP while bringing in Hide and other ACT MPs. This tactical voting for conservative voters in this one electorate isn’t available in most other electorates.

The concept of allocating seats to a party on 2% while not allocating seats to a party which won 3% also retains the issue of favouring parties with geographically concentrated representation. MMP also encourages independent MPs (such as those who have left a major party) to establish a party which they dominate rather than remain as an independent. This has resulted in high personal votes for one MP resulting in a group of other MPs, totally unknown to the public, on no other basis but their loyalty to their party leader, as was seen with United Future New Zealand in 2002.

The other major issue with New Zealand’s system is the distinction between list MPs and electorate MPs, and the problem inherent in all party-list electoral systems. While MPs in the Irish and Tasmanian versions of PR-STV, and in most single-member electorate systems, may get elected on the back of their party endorsement, local members generally are required to have a minimum level of community support, competence and effectiveness to win. In contrast, the New Zealand system, as well as bodies such as the New South Wales Legislative Council, sees people with no respect for the electorate’s opinion of them.

The single-member electorates also mean that, for Maori electorates, representation remains restricted to the First-Past-the-Post era, with the Maori Party on track to win all seven seats, despite remaining below 50% support in the Maori community.

So what’s the alternative? An Irish-style PR system, perhaps including one STV electorate covering Maori voters, would see a much better distribution of local members of Parliament and would remove the anomaly of MPs elected with no popular basis. MPs like Peter Dunne would remain a highly popular independent in one electorate, rather than becoming the leader of a large party bloc despite being the only name known to any New Zealanders. However, it appears that there is little appetite in New Zealand for further improvements, with most pro-PR campaigners fighting to defend MMP from those, particularly in the National Party and the business community, pushing for a return to first-past-the-post.