Via IFES’ ElectionGuide, the Pacific island of Nauru will hold a February 27 referendum on whether to switch to direct presidential elections. Currently, Nauru’s president is essentially a prime minister, elected by and subject to no-confidence votes of its Parliament.
Why the change? According to the Australia Network News, the idea is to shake a history of cabinet instability:
Nauru’s Justice, Health and Sports minister, Mathew Batsiua, says it is hoped that change will help end the volatile nature of Nauru’s politics, which has resulted in around 36 changes of government since independance [sic] in 1968.
“[I] think this is the right structure to gain political stability, and I think that’s what Nauru badly needs,” he said.
I’m no expert on Nauruan political history, but 36 governments in four decades is a lot.1 According to the CIA Factbook, the most recent change of government was in December 2007, when a no-confidence vote sacked President Scotty.
Unfortunately for those who support reform as a remedy for cabinet instability, the proposal for direct elections does not come with a complementary proposal to end presidential dependence on legislative confidence. Features of the Nauruan electoral system, moreover, may magnify the country’s stated troubles.
It may be a stretch to emphasize institutions in this 21,000-person country with its 18-member legislature, but some points are worth note. First, Nauru uses a candidate-based and effectively majoritarian electoral system: a modified Borda Count in two- and four-seat districts (seven and two, respectively) that weights voters’ higher rankings more heavily than lower ones. Second, elections in Nauru are non-partisan, whether by law or by fact. Notwithstanding, from what I can tell, candidates tend to line up behind one of two de facto party leaders,2 an outcome consistent with the country’s electoral rules. It’s also an outcome that may resonate with PR advocates who cite the tendency for ‘pendulum swings of government’ in parliamentary systems with majoritarian electoral formulas.
By contrast, the Netherlands Antilles, only twice Nauru’s population, (population 227,000) uses an open-list PR system within a parliamentary framework. If we accept that Nauru has two de facto parties, there is much more diversity in the Netherlands Antilles’ party system, an outcome consistent with its low-barrier voting system. My impression is that cabinets there generally have been more stable. Of course, the Netherlands Antilles are made of several islands – Nauru is not – and the Netherlands Antilles enjoy less sovereignty than Nauru, which could imply that the former’s domestic politics carry lower stakes. On the other hand, Nauru’s economy appears much more dependent on natural resource extraction, a fact that often coincides with extensive patronage and, as a result, governments less apt to change.
The Republic of Nauru has been quite a laboratory for institutional experimentation. It is one of three countries I know to use the Borda Count in national elections. The other two are Kiribati and Slovenia (for two ethnic minority seats in its legislature). According to Ben Reilly, the adoption of Borda Count appears to have been a reaction to “complicated” preference transfers under the previous alternative vote/IRV system.3 Its weighting of higher rankings is also novel. As Reilly notes, this modifies Borda’s original plan for a system in which preferences decreased in weight by a constant value. Within the context of the low-magnitude multi-member district system, then, this system has a strong MNTV (a.k.a. “winner-take-all at-large”) flavor.
The direct election proposal is one of four constitutional amendments coming before voters in the country’s first ever referendum. Regardless of the underlying drivers of Nauruan cabinet instability, the proposed institutional remedy is unlikely to address it because it would not end executive dependence on parliamentary confidence (see Article 24 of the aforelinked). Nor is any reform proposed that would address the essentially majoritarian character of elections in Nauru.
I do not have the mathematical competence to prove it but I suspect there is a lower limit on population size for an effective sovereign state. Politics in the Pacific Forum states, most notably Fiji, has not been spectacularly stable. Nauru, which is among the smallest of them, seems to rival Fiji for governmental instability. Dependence on an external power may be the most important difference between Netherlands Antilles and Nauru.
Grand Chief Sir Michael Somare, the Prime Minister of Papua-New Guinea (the mix of titles emphasises one of the peculiarities of Pacific politics) has speculated that the Melanesian Spearhead states have all suffered degrees of instability and related that to the personalised nature of Melanesian political relationships.
Seed planted by Alan — 17 February 2010 @ 01:50
Actually, the Netherlands Antilles has a population of about 200,000 – ten times as large as Nauru’s.
Also, while it’s true that their unicameral Parliament, the 22-seat Staten, is elected by PR, there’s only one island in which a reasonably proportional outcome is attainable, namely Curacao, which holds 14 seats; (Dutch) Sint Maarten and Bonaire only have three seats each, while Saba and Sint Eustatius are both single-member constituencies.
Moreover, the electoral system requires parties to win a full Hare quota in order to take part in the proportional allocation of seats in a multi-member constituency. Thus, in last January’s federal election in St. Maarten, the National Alliance won all three seats there with 60.7% of the vote, since both the Democratic Party (31.2%) and the People’s Progressive Alliance (8.1%) fell short of the one-third threshold in that island.
Incidentally, each island has developed its own, distinct party system; the larger island, Curacao has a decidedly fractious party system (not surprisingly, since the threshold there amounts to 7.1% of the vote). However, none of this really matters much anymore, since the federation of the Netherlands Antilles is set to be dissolved next October 10: on 10/10/10, Curacao and Sint Maarten will become autonomous countries within the Kingdom of the Netherlands – the same status enjoyed by Aruba since 1986 – while the remaining islands will be integrated to the Netherlands as special municipalities.
Seed planted by Manuel Alvarez-Rivera — 17 February 2010 @ 18:52
Thank you, Manuel. I have amended the population of the Netherlands Antilles accordingly.
My speculation about OLPR there was related to its effect on the number of parties, and the effect of this on coalition politics, rather than any effect of an electoral system on proportionality. One hypothesis is that NA’s (relatively) permissive electoral system allows for each island to have its own party system.
Seed planted by Jack — 17 February 2010 @ 19:45
Is this proposal essentially a repeat of the (disastrous) Israeli experiment with what I call an “elected prime ministerial” regime?
That is, directly elected head of government, who (along with the cabinet) survives in office only so long as he or she has the confidence of a legislative majority.
In Israel, the result was worsened executive-legislative relations. (See chapter 6 of Samuels and Shugart, forthcoming.)
Seed planted by MSS — 21 February 2010 @ 23:48
The draft constitution, if the referendum passes, can be found here. The presidential election follows the parliamentary election and only MPs may be candidates. The parliament stands dissolved automatically if it passes a vote of no confidence.
Seed planted by Alan — 22 February 2010 @ 14:28
Didn’t Kenya formerly operate a similar system – the winning presidential candidate had to also win a parliamentary constituency, or else forfeit?
I understand Tasmania has a similar rule: the directly-elected mayor must also win a council seat. If a mayoral by-election is held, only sitting councillors may nominate.
I would propose an alternative way of combing a presidency with a premiership – modelled on Australia’s sole respectably-sized quasi-presidential jurisdiction, the Brisbane City Council. I would have, alongside the 20 or 100 or 300 or whatever ordinary “quota members” (so named because they are elected by PR quota, or represent a quota of population within an electoral division, or both), one extra “member at large” chosen by majority vote of the whole.
The member at large would not automatically become mayor/ governor/ president – that position would be voted on by the council/ legislature – but if no other candidate got an absolute majority on the first ballot or two, the member at large would win by default. (If s/he declined to stand, a runoff by plurality would decide). IOW, the member at large could only be beaten by a candidate with absolute majority support. If there is a disciplined majority party or tight coalition, they can elect their leader as mayor/ governor/ president, as per usual Westminster system, avoiding a Washington-style gridlock (or, indeed, Brisbane CC for Campbell Newman’s first four-year term). On the other hand, if the council/ legislature has numerous independents and/or minor parties, Weimar-style instability is avoided by the de Gaullian solution – a chief executive chosen by direct majority vote. (“Either a door is open, or it is shut”, the General said, pushing for direct Presidential runoff elections).
Seed planted by Tom Round — 23 February 2010 @ 00:53
I look forward to our hosts forthcoming article Morning After Shotgun Marriage elections in Nauru.
Seed planted by Alan — 23 February 2010 @ 01:20
More seriously, the Nauru proposals are not identical with the Israeli experience. Samuels and Shugart note:
If the proposals are adopted, a vote of confidence against the president will automatically dissolve the assembly as in Israel, but any other presidential vacancy will also entail a new legislative election. Once a new assembly is elected, the MPs must then nominate 2 or 3 of their number to run in the subsequent presidential election.
There is also an interesting body called the Council of State which becomes a collective acting presidency when the president loses a vote of confidence or is removed under the new leadership code. The members are the chief justice, speaker and chief secretary (permanent head of the civil service).
Seed planted by Alan — 23 February 2010 @ 05:21
It looks like the changes will not be adopted.
Seed planted by Jack — 02 March 2010 @ 15:37
- Hedley Thomas and Michael McKenna, ‘Nauru instability “Aussie phosphate plot”,’ The Australian (Monday 25 October 2010).
A dead heat result with 18 seats? Who’d'a thunk. Fellas. Word of wisdom. Sweden 1973 (175-175) was unexpected. Spain 1989 (also 175-175, oddly enough) was also unexpected. Even Australia 2010 (75-74) was unexpected. 150 seats is a big enough coin that you don’t expect to land on its edge. 18 seats is small enough for ties to be entirely predictable. Go ACT or Tasmania – 17 or 19. As Shakespeare said, there is magic in odd numbers…
Seed planted by Tom Round — 27 October 2010 @ 06:08
“I do not have the mathematical competence to prove it but I suspect there is a lower limit on population size for an effective sovereign state. ”
I suspect the same thing and would like to see this looked at more. But there also seems to be an international trend towards microstates.
I also suspect that there is a version of this law that applies to subordinate jurisdictions.
Essentially, once a jurisdiction has a particularly small relative population combined with similar jurisdictions, it becomes effectively “governed”, if at all, by some interest other than the nominal government. The nominal government just doesn’t have enough heft to secure its monopoly.
Seed planted by Ed — 27 October 2010 @ 14:53
This planting needs revision. The Dowdall rules in Nauru do not make the Borda system there more like MNTV, but instead more like SNTV. This is because a lower ranking carries far less value than a higher ranking. The point is probably trivial, though, because the district magnitudes are so low.
Seed planted by Jack — 08 November 2010 @ 23:48