Wednesday, March 30, 2016

An early Japanese election?

In recent days, there has been speculation that Japanese Prime Minister Shinzo Abe will call an early election for the House of Representatives of Japan, after delaying an increase in the sales tax. This would be Abe's second usage of this tactic; he called an early election in December 2014 for the same reason, which he won easily.

Abe currently has a commanding majority in the House of Representatives, which is the house of the bicameral Japanese Diet that has the power of confidence over the government. His Liberal Democratic Party has 290 seats out of 475. If you add the votes of Abe's coalition partner, the Komeito Party (a Buddhist political organisation), the government has 325 seats, or slightly more than a two-thirds majority.

In the House of Councillors, the upper house which has the power to block legislation (but can be overridden by two-thirds of the lower house), the Liberal Democrats have 116 seats (with 121 needed for a majority) and the government coalition has 126. Even if they did not have a majority in this house, Abe could still pass his legislation with his two-thirds in the Representatives.

Like Australia, the Councillors have fixed six-year terms, and half of the members are elected every three years. Unlike Australia, however, the frequency of upper house elections does not match up with the maximum length of lower house elections. While the Representatives can be dissolved at any time, the maximum term is four years.

Also unlike Australia, the Councillors cannot be dissolved as a deadlock-breaking mechanism; while the 2005 lower house election was held to break a deadlock over privatising the postal service, the goal was to win two-thirds in the lower house.

This discrepancy has meant that Representatives and Councillors elections have rarely happened at the same time. It has only happened once in the history of the post-war Constitution; in 1980, when the Liberal Democratic government was defeated on a vote of no-confidence at the same time as the Councillors election. The government won a majority in both houses, including an unexpectedly large majority in the Representatives.

The last Councillors election was in 2013, which means that an election is scheduled for this year; July, to be precise. As a result, if an early election happens this year, it would make sense for it to be in July, to reduce administration costs and to keep turnout up.

So, that being the case, would an election make sense?

Arguments for an election

It would avoid a midterm election. Midterm elections can be dangerous for governments, as they allow voters to express dissatisfaction with the government without necessarily throwing it out. While this can be embarrassing for governments if the elections are for subnational districts, it is dangerous if the election is for a house with direct control over the government.

The effects of a general election, at which the fate of the government would be at stake, would likely give the election winner (likely to be Abe's Liberal Democrats) a boost in the upper house elections, perhaps even giving the Liberal Democrats an absolute majority.

This is an especially important reason, given that a recent redistribution has increased the number of single-member districts in the Councillors. Councillors elections take place using a mixed-member majoritarian system; 78 members are elected in districts using the single non-transferable vote, and 48 members are elected using party-list proportional representation in one nationwide constituency, with open lists.

The redistribution has reduced the number of two-member districts, and increased the number of single-member districts. Two-member districts are very hard to win two seats in, as, in a hypothetical race between three candidates (one from the opposition, two from the government), the government needs two-thirds of the vote to win both seats, assuming equal vote division. This normally means that each party only nominates one candidate, and the two seats go to the top two parties.

In a single-member district, however, the largest party only needs to win the most votes to sweep all of the seats. As a result, in the districts that have become single-member, the largest party will win it all.

For that reason, a drop in support for the Liberal Democrats could lead to a more dramatic loss in seats. Abe cannot afford such a drop; losing control of the Councillors would complicate the government process, and would likely weaken his strength within the Liberal Democrats.

It would strengthen Abe's leadership. Japan has a tendency towards revolving-door Prime Ministers; the 2009-2012 Democrat term in government had three Prime Ministers, and the preceding 2005-2009 Liberal Democrat government had four Prime Ministers. Abe's three-year Prime Ministership looks exceptionally long.

Perhaps this speculation of an election has come about because of unease within the Liberal Democrats as Abe's leadership. An election is a powerful tool for party unity; no faction would dare undermine the leader during an election, an election could allow Abe to attempt to get rid of his factional enemies, and an election would provide Abe with a personal mandate, which would make him harder to get rid of.

Arguments against an election

Unusual opposition unity. From the formation of the Liberal Democrats in 1955, until about 2000, Japan had a 'dominant-party' system, with the Liberal Democrats winning around half the votes and somewhat more than half of the seats, and the remaining seats going to the opposition parties; the Socialists, Communists, Komeito, and the centre-left Democratic Socialists. The Liberal Democratic government was dominant, because there was no alternative government; while the Socialists were the largest opposition party, they did not have enough seats to present the possibility of forming the government.

In 1993, the first non-Liberal Democratic government since the formation of the party was formed. It was a broad coalition of the traditional opposition Socialists, Komeito and Democratic Socialists, as well as a number of parties formed by Liberal Democratic dissidents. While it did not last especially long, one of its key achievements was the replacement of Japan's single non-transferable vote in 3-5 member districts with a mixed-member majoritarian system, in which 300 members would be elected in single-member districts and 200 (later 180) members elected in 11 multi-member districts using party-list proportional representation.

This electoral system was meant to create a two-party system, and encourage more competitive elections, as the opposition parties would be encouraged to merge by the majoritarian effects of the new electoral system. While it took a while, in 1998 the Democratic Party of Japan was formed, as a centre to centre-left party comprising social democrats, conservatives and liberals. The Komeito Party formed a coalition with the Liberal Democrats, and what remained of the Socialists renamed themselves the Social Democratic Party. The Communists remained independent.

At the DPJ's first election in 2000,  the top two parties received 53.5% of the party list vote and 75% of the seats. This jumped to 72% of the vote and 86% of the seats in 2003, and was 69% of the vote and 89% of the seats by the time of the 2009 election, which was the year of the DPJ landslide. 

However, the DPJ's time in government exposed internal splits. The number of government MPs collapsed from 308 to 230, with those that left joining the anti-nuclear People's Life, Tax Cuts and Kizuna parties to the left and the nationalist Japan Restoration Party to the right (a much rarer move). While the anti-nuclear party did not perform so well at the 2012 election, the nationalists polled strongly, reducing the support for the top two parties to 43% of the vote and 73% of the seats. 

Since this election, there have been more party splits, with the Restoration Party splitting into two. This has made presenting an alternative government to the Liberal Democrats a difficult task, as there is little chance of them winning a majority.

However, in recent months the major nationalist party, the Japan Innovation Party, has decided to merge with the Democratic Party of Japan. The party formed is called the Democratic Party, and will be led by DPJ leader Katsuya Okada. It will also be joined by members of the tiny Visions for Reform party, which will give it nearly 100 seats in the Representatives and 71% of non-government seats. This would be enough to present a convincing alternative government.

It is worth noting that such an alliance may not necessarily be long-lasting. And this is a key disadvantage of an early election; it does not give this new party time to have factional conflicts, and potentially split. Abe would be going to an election at the time at which the opposition would be most unified, a courageous decision to be sure.

The stakes would be very high. As stated before, an early election would see Abe facing a more unified opposition. If the election is just for the Councillors, he could afford to lose this and continue to pass legislation. However, if his popularity wanes before a Representatives and Councillors election, the government could end up losing both its Councillors majority and its two-thirds Representatives majority, which would almost certainly lead to the end of Abe's party leadership.

So, what will happen?

In my view, it would not make sense for Shinzo Abe to call such an early election. Such an election would be a high-stakes poll that, if it goes poorly, could lead to Abe losing control of the Diet, which would almost certainly mean the end of his Prime Ministership. As for the consumption tax, an increase could be delayed until after the Councillors election, which could give him a boost.

However, speculation of an election is likely to stick around for longer, since Abe may need the threat of an election to keep members of his party in line, as talk of a consumption tax increase leads to a drop in his approval ratings. For this reason, if Abe is sensible, there will be no election, but there will be talk about it.

Thursday, March 24, 2016

Why does the Thai junta want MMP?.

Thailand is currently under the rule of a military junta, following a coup in 2014 against the government of Yingluck Shintawara. The coup is nothing especially unusual for Thailand, which has had 19 attempted coups, 13 of which have been successful.

The military government has said that they intend to return Thailand to democracy eventually: although they have not provided an exact date, it is expected that it will happen later this year, or next year. One of the key tasks that the junta has stated needs to take place before democracy is introduced is writing a new constitution.

Since the constitutional monarchy was introduced in 1932, Thailand has had seventeen constitutions and charters. Charters are documents set up to provide a basis to temporary military governments, and normally have been used for two or three years at a time. Constitutions have been more permanent, but none have lasted especially long.

The current constitution was drawn up and promulgated in 2007, following a coup in 2006. It provides, like all the other democratic constitutions, for a parliamentary system of government with a constitutional monarchy. A bicameral system was established, as had existed in the previous Constitution; however, the Senate was not able to remove ministers, and half of it was appointed. The House of Representatives was the premier house; it was directly elected, through a mixed-member majoritarian electoral system. Oddly, the exact provisions of the electoral law are enshrined in the Constitution.

Formally, the Prime Minister and the remainder of the Cabinet are appointed by the King; however, a no-confidence motion from the House can remove Ministers. This is similar to other constitutional monarchies, such as Australia and the United Kingdom. However, for various reasons (including the popularity of the King, and lese-majeste laws banning criticism of him) the King has had a more influential role than other constitutional monarchies.

The key factor in the two recent coups has been the governments of Thaksin Shinawatra and his sucessor and sister Yingluck Shinawatra (many female leaders in Asia have been related to former male leaders, such as Prime Minister Indira Gandhi of India or President Park Geun-hye).

Thaksin came to power following the Asian Financial Crisis of the late 1990s. He was an influential businessman, and formed a political party called Thai Rak Thai (Thais Love Thais). At the 2001 election, his party won 248 seats in the 500 member House of Representatives, a near majority, off 40.6% of the vote. This result was part of a growing trend of Thai politics towards consolidation of the party system, as the below graph shows.

As you can see, the 2001 election represented the largest share of the votes for the top two parties since vote shares became available, and probably before that. This is off the back of three subsequent increases.

Exactly why this is the case is unclear. However, Thailand's usage of a different electoral system may account for some of this. Just before the 2001 election, Thailand switched from a pure multiple non-transferable vote system to a mixed-member majoritarian system, with first-past-the-post used in the local districts. 

Shugart and Carey (1995) provides a rank-ordering of electoral systems based on the importance of the personal vote. The issue with this is that the rank-ordering places substantial emphasis on how parties nominate candidates and the ease of getting on the ballot, and information on this factor is unavailable. 

As a result, the switch to MMM would be likely to reduce the incentive to cultivate a personal vote. This could be expected to lead to party labels becoming more important in elections, and thus could be expected to lead to stronger, more disciplined parties. I do not have many cases to directly compare the Thai experience to, seeing as the MNTV system is relatively rare. However, the replacement of Taiwan's MMM system, which used the single non-transferable vote in multi-member districts for the local tier, with a MMM system that used single-member plurality in the local tier led to what was a multi-party system consolidating into a two-party system.

After the election, the Thai Rak Thai party merged with the New Aspiration Party, which had won 36 seats. This gave Thai Rak Thai a majority in the House of Representatives. Thaksin was thus able to implement a populist electoral platform, introducing subsidies for health care and petrol, constructing large amounts of new infrastructure, and low-interest loans for farmers.

Thaskin's policies proved popular, and he won a landslide victory at the 2005 election. His party won 377 seats, and approximately 62% of the vote. Thai Rak Thai dominated in the north and northeast of the country; this area is populated with poorer farmers, who would be most responsive to Thaskin's handouts. The infrastructure projects funded by his government allowed Thai Rak Thai to poll strongly in wealthier Bangkok. However, the south of the country was still dominated by the opposition centre-right Democratic Party. This region gave the Democrats half of their seats in the single-member districts alone.

The Thai Rak Thai majority was the first for any party in Thai history. The election also represented a peak of political consolidation, as 85% of votes were cast for the top two parties. Thaskin had been the first elected Thai prime minister to serve a full term in office.

However, all was not smooth for Thaskin. He faced allegations of corruption after his family sold shares in a telecommunications company following legislative changes that allowed the shares to be sold to a foreign owner. He denied claims of corruption, but, after protests against his government began, he decided to call an election in order to test his support.

While the election was called, the major opposition parties decided to boycott it, claiming that they were unfair and that Thailand needed constitutional reform before a fair election could be held. As a result, the elections were only contested by the Thai Rak Thai party, and a number of small unidentified parties.

In normal cases, election boycotts involve electors not turning up at the polls at all. However, Thailand gives voters the option of voting for 'no candidate' by ticking a box at the bottom of their ballot paper. In cases where only one candidate ran for election, the election would still be run, and a candidate would need support from 20% of registered voters. This would not be such a dramatic threshold for Thai Rak Thai MPs to win in most parts of the country. However, in the south, there were seven constituencies where Thai Rak Thai did not win 20% of the vote, and a further fourteen where the Thai Rak Thai candidate won less than 25%.

In these constituencies, the election boycott could be expected to lead to the Thai Rak Thai candidate not reaching the threshold. On election day, however, Thai Rak Thai lost forty seats, which had to be filled through by-elections. Even after these by-elections had been held, in which candidates alleged to be Thai Rak Thai 'dummies' filed in most seats in order to avoid the 20% absolute support threshold, another fourteen by-elections were needed.

However, at this point, the Constitutional Court upheld a complaint by the opposition that the elections were unconstitutional. Thaskin's cabinet decided not to contest this, and called elections for October of that year; just one month away. Democracy in Thailand would not last that long.

In the night of the 21st of September, military forces entered Bangkok and took control over the institutions of government. Thaskin was in New York at the time, intending to give a speech to the United Nations. The military government introduced new, strict security measures, but promised that democracy would be re-introduced 'in a year'.

The junta did, actually, keep their promise. After a year of heavily restricted political activity, a referendum was held in August 2007 on a new constitution. The patterns of voting followed traditional Thai patterns, with the South voting yes, and the North voting mostly no. 58% voted Yes, thus meaning parliamentary elections could be held later that year. Some changes were made to the electoral law; multi-member constituencies, elected using the multiple non-transferable vote, were used for some of the district tier seats. The number of party-list seats was reduced to eighty, and they were elected in ten-member districts.

Elections to the Representatives were scheduled for December. Thai Rak Thai had been banned, along with Thaskin from the country, so those who supported it needed a new political party. The new party formed was called the People Power Party, and was led by former Governor of Bangkok Samak Sundaravej. The Democrats were not dissolved, and contested the election. No other major forces entered the race, thus meaning that, to some extent, Thailand was back to where it started.

In the end, People Power won the election, securing 233 seats in the 480 member assembly; slightly short of a majority, but it would be expected that they would form the government. The Democrats won 165 seats, the conservative Thai Nation Party won 37 seats, and the moderate Motherland Party won 24. No other parties won more than ten seats.

What is interesting about this election is that the party vote was actually very close. Statistics conflict, but the People Power-Democrat gap was very narrow either way. The large lead for People Power in the seat totals was due to their substantial lead in the district seats, where the People Power vote was mostly the same; however, the opposition was more heavily split. There were only eighty proportional seats. This will become important later.

People Power formed the government, and Sundaravej became Prime Minister. However, he was opposed by the military, which still held substantial sway in Thai politics. In September 2008, he was found guilty of accepting payment from the private sector while in the office of Prime Minister: he hosted a number of cooking shows, for which he was paid. He was replaced by Somchai Wongsawat, a relation of Thaskin. 

However, his term was also short-lived. In December 2008, the People Power Party was banned for vote-buying and other corruption. At the time, there were massive public protests against the government, and protests for the government. Following the banning, the coalition partners of the People Power Party and some dissident People Power MPs formed a coalition with the Democrats, allowing Democrat leader Abhisit Vejjajjva to become Prime Minister.

Vejjajjva's government was opposed by the newly formed Pheu Thai Party, which was led by former Prime Minister Wongsawat. This group had the support of the 'National Front for Democracy against Dictatorship', also known as the Red Shirts, which led protests against the Vejjajjva government calling for new elections.

One of the actions of the Vejjajjva government was a change to the electoral law. The number of party-list proportional representation seats was increased to 125, out of a 500 member legislature. Allocation was also changed to a nationwide level. The district tier was changed to single-member plurality only. These changes could be expected to advantage the Democratic Party, which had been substantially stronger in the party-list tier than in the districts.

The House was dissolved in 2011, and elections were called for July. Pheu Thai were led into this election by Yingluck Shinawatra, the sister of Thaskin. Vejjajva stayed as Democrat leader. Thai polls, as rare as they were, gave Pheu Thai a comfortable lead over the Democrats.

On the day, Pheu Thai won 265 seats in the 500 member House, a fairly comfortable majority. They won 48% of the party list vote, fairly close to a majority. The Democrats won 159 seats, and 39.6% of the party list vote; the electoral system still disadvantaged them, as this was only 31.8% of the seats. This was probably due to them only getting 32.3% of the district vote.

Oddly, the smaller parties were less represented in the proportional tier than in the districts. The Thai Pride Party, a populist party formed by former Thai Rak Thai members, received 11% of the vote in the single-member districts, and 29 out of 375 seats. However, they won only 3.9% in the party-list tier, and five seats. The National Development Party won 15 seats in the districts and 4.6% of the district vote, but only 2.8% of the party-list vote.

Following the election, Yingluck was appointed Prime Minister, after forming a (seemingly unnecessary) coalition with the Thai Pride and National Development parties. Her government followed similar policies to the previous Thai Rak Thai/People Power governments, and, unsurprisingly, faced the same protests that those same governments had faced.

One of the most controversial initiatives of Yingluck's government was an attempted amendment to the 2007 constitution. As constitutions go, the Thai one was at that time was relatively easy to amend. It required only an absolute majority vote of both houses, and a public hearing. Yingluck wanted to bring back a fully elected Senate, and passed it through the normal process. However, the Constitutional Court found that the reform was unconstitutional, on the basis that it had been changed while it was passing through the legislature.

2014 saw more political violence, as activists from both sides of politics sparred over the government's record. Political figures from both sides of the divide were killed and injured, and Yingluck, desperate to regain control over the situation, called an election in February 2014. However, this did not go so well. As in 2007, the opposition boycotted the election, but this time the security situation had deteriorated to the extent that the Electoral Commission was unable to hold the election at all in a substantial number of constituencies, and that in some seats, candidates did not nominate at all.

The election was invalidated by the Constitutional Court, however, as it had not been held on the same day in all areas. Weeks later, Yingluck was stripped of her office as Prime Minister over the transfer of a military official, and was replaced by the Deputy Prime Minister.

At this point, the military decided to take control over the government. On 22 May 2014, the military held meetings between the leaders of  the civilian parties. When the talks failed, General Prayut Chanocha informed the meeting that he had to seize power, and ordered the military to once again take control over government.

Once again, the military said that their control over government would be purely temporary, and that they would restore democracy at some point. They introduced a new interim constitution, which starts by saying that Thailand has a "democratic regime of government" before enshrining an appointed legislature and giving the poetically named "National Council for Peace and Order", the executive under the new regime, the power to decree anything they wished. Decrees would immediately become constitutional upon introduction.

So, what does this have to do with MMP?

For those of you who are desperate to hear about MMP, this is the part where it becomes important. The Constitution Drafting Committee has been working on a new constitution since the coup, and one of the things that they have suggested is a new electoral system for Thailand.

The option that has been most mentioned is a mixed member proportional representation system. Now, there is nothing unusual about this electoral system, and it has been proposed for many nations recently during electoral reform processes. However, your author cannot help but be a bit suspicious of the proposal when it comes from a military junta. So, the question is, what would be in it for the military if the MMP proposal was introduced?

The military are clearly no fans of the Thaskinite governments. As a result, it would make sense for the military to introduce an electoral system that disadvantaged these parties. While the Democrats deny such a connection, it is generally considered that they are closer to the military than the Thaskinites. It would also be an asset to the military government if the new electoral system would advantage the Democrats.

Below are the results of the 2011 election under various forms of MMP. From left to right; all parties with 5% of the vote allowed to win proportional representation seats, all parties with 5% or one constituency seat, all parties with 3%, and all parties with 3% or one constituency seat. The final column is the actual results of the election.

Click on the image for a larger size

My careful readers will note that the total number of seats is higher under all of the MMP systems. This is because of a feature of MMP called 'overhang'. Under most versions of MMP, a party which wins more seats in the districts than it is entitled to by the party-list calculation keeps those district seats, and the legislature is enlarged to compensate. With a 5% threshold, only Pheu Thai and the Democrats are entitled to the proportional representation seats, and therefore all 56 districts won by parties outside these two are overhang.

There is one consistent trend, if you look at the percentages. The Democrats always have a higher percentage of seats under a MMP system, because of their higher share of the party-list proportional representation vote. This gain mostly comes from the smaller parties, which are no longer able to boost their seat totals through the PR tier.

It is evident that the Democrats perform better in a nationwide ballot, so it is evidently clear that increasing the importance of the party-list ballots through mixed-member proportional representation would be good for them.

Another key advantage of MMP for the military is a likely increase in political fragmentation, and a weakening of the party system. MMP would make it easier for small parties with nationwide support to enter the House of Representatives. This would likely lead to a fragmented party system, as there would be less incentive for parties to merge.

An increase in political fragmentation would be advantageous to the military for one key reason; it would weaken the influence of civilian governments against the military, as they would be less stable without regular parliamentary majorities. It could be said that this is a good thing, as it means that the populists need to look for consensus more. However, it is clear what the military gets out of the arrangement.

All of this does not mean that mixed-member proportional representation would not necessarily be a good idea for Thailand. In my view, it would be a good idea to encourage consensus in what is evidently a heavily divided political culture. However, it is also evidently clear that the advantages for the military are very clear in implementing MMP, and the specific proposal will warrant very close examination in the context of the other constitutional features.