The crisis of the 2021 Samoan election


Last month’s election in the small Pacific Island nation and the constitutional crisis that continues into this week has me asking: Was Samoa ever a democracy? By any intuitive account, yes. It has free elections for a parliament from which ministers are appointed and a head of state is elected, and the Samoan constitution grants several civil liberties, namely freedom of expression.

For some people, namely political scientists, these intuitive accounts are not enough. This post will take a more analytical approach, comparing a number of democracy indices, the background to the election campaign, analysis of the results, the myth of ‘dominant-party’ democracy, and remarking on the aftermath.


Editorial note: This post was published prior to the resolution of the various court battles relating to the election in May 2021.


Due to the length of this post and the many ideas it raises, I’ve included a table of contents for your reading comfort.

  1. Common democracy indices and their problems
  2. The Joy of DD, and Samoan “democracy”
  3. The conduct of the 2021 election campaign
  4. The results and the downsides of first-past-the-post
  5. The aftermath and the myth of ‘dominant-party democracy’
  6. What now for Samoa?
  7. Conclusions

Common democracy indices and their problems

Back to top

The Economist Intelligence Unit’s Democracy Index and Freedom House’s Freedom in the World report are among the more popular indices used by the public, but for answering the question above it is useless. The most recent edition of The Economist‘s Index does not list Samoa at all and it does not give enough data for replication.

Freedom in the World‘s 2020 edition gives an overall “freedom” score from based on two questionnaires asking questions about economic and civil rights, on which Samoa scored 8.1 out of 10. In their report, Freedom House’s experts said this:

Is there a realistic opportunity for the opposition to increase its support or gain power through elections?

There are no obvious obstacles that prevent the opposition from increasing its support and gaining power through elections. However, the ruling [Human Rights Protection Party] has been in power since 1988 and has developed an effective campaign machinery during its incumbency, raising concerns about whether its long stay in power is due to the party’s popularity or features of the electoral system that may put the opposition at a disadvantage.

Freedom House. Freedom in the World 2020: Samoa. (Emphasis added)

Freedom House, despite admitting that Samoa’s democratic consolidation is unclear, proceed to give glowing three-out-of-four scores on each question, each with an answer that read more ambivalently than the scores they are attached to, before ultimately concluding that Samoa is free – the best category on the scale.

Freedom House’s methodology is inconsistent as it is theoretically possible for people to precisely follow the methodology document for Freedom in the World and end up with a different result than that which Freedom House came to.

For example, take the question on trade unionism: “Is there freedom for trade unions and similar professional or labor organizations?“. Australia is given a score of 4-out-of-4, the “freest” option, as trade unions are indeed ‘free’ to exist and organise. However, as Freedom House notes, the right to strike – arguably the biggest bargaining chip available to a trade union – is taken by the courts to be a limited and revocable “privilege” conferred by the Fair Work Act. Contrast this situation with France, where “union action” is an expansive constitutionally-protected right, which Freedom House also gives the best possible score with virtually no comment.

In the above example, an independent scholar following Freedom House’s methodology document could argue that Australia deserves a lower score on this metric than France based on this notable difference between a ‘right’ and a ‘privilege’. Or they could argue in agreement with Freedom House’s determination. The methodology document only provides more questions to consider when giving a score.

The Joy of DD, and Samoan “democracy”

Back to top

Przeworski, Alvarez, Cheibub and Limongi (PACL)’s typology of democracies and dictatorships (DD) has a different approach. DD takes a minimalist observational approach that simply observes whether various institutions in the country are elected and (crucially) whether those elections matter where the opposition wins.

DD holds that a country is a democracy where it has free elections, the elections are respected, and there is an example of an opposition party winning and taking office. Otherwise, the country is run by a dictatorship. It does not have a scale of ‘more’ or ‘less’ democratic, it is a simple binary which can be further divided into various institutional forms of dictatorships and democracies. Because of its simple rules, it is highly replicable and only requires a basic, Wikipedia-level familiarity with a country’s electoral system and history to get a clear answer.

A regime as classified as a dictatorship during a particular year if at least one of these conditions holds:

Rule 1. “Executive Selection.” The Chief Executive is not elected.

Rule 2. “Legislative Selection.” The Legislature is not elected.

Rule 3. “Party.” There is no more than one party. Specifically, this rule applies if (1) there were no parties, or (2) there was only one party, or (3) the current term in office ended in the establishment of a non-party or one-party rule, or (4) the incumbents unconstitutionally closed the legislature and rewrote the rules in their favour.

Rule 4. “Type II Error”. A regime passes the previous three rules, the incumbents will have or already have continuously held office for more than two terms or without being elected for any duration, and until today, or the time when they were overthrown, they have not lost an election. [Also known as the ‘alternation rule’]

M E Alvarez et. al. 1996. “Classifying political regimes” in Studies in Comparative International Development no. 31: page 14

According to DD’s rules, Samoa fails on the fourth rule because although there are free elections for the legislature (which, in turn, elect the prime minister and head of state), there has not been a change of government at an election under the same rules that elected the current government since at least 1982. This means that, at least according to DD’s observational approach, Samoa is a dictatorship because a democratic transition of power has never been observed.

The Human Rights Protection Party (HRPP) has, until last month’s elections, won at least a plurality of the seats in the Fono (parliament) in every election since 1982 and majorities in every election since 2006. It has been led by Tuilaepa Aiono Sailele Malielegaoi since 1998. Tuilaepa is the second-longest serving prime minister in the world, behind Cambodia’s dictator Hun Sen.

A common criticism of DD is that it over-classifies countries with stable democratic governments as dictatorships, but I argue that DD’s classification of Samoa as a dictatorship has only been confirmed by the recent constitutional crisis. DD does not concern itself with normative assumptions of whether a regime would democratically transition if it hypothetically lost an election, it only concerns itself with can actually be observed.[note]

Even Freedom House, in the quote in the previous section, admits that it is possible “features of the electoral system that may put the opposition at a disadvantage” but they gloss over these potential flaws — on the normative assumption that Tuilaepa would peacefully stand down after 23 years in power — and give them a glowing assessment.

The conduct of the 2021 election campaign

Back to top

From the start of the campaign in late 2020, things looked like they were going to get very ugly and there were no signs from the outset that Tuilaepa would concede defeat if the situation arose.

The 2021 election was to be the first under the new boundaries and a new Electoral Act drawn up by the HRPP-controlled parliament, which will be important to understanding the result later.

In September 2020, the then-Deputy Prime Minister – Fiame Naomi Mataʻafa – was expelled from cabinet and from the HRPP after she refused to support the Land and Titles Bill and its related constitutional amendments.

The controversial laws introduced ouster clauses which removed the supervisory jurisdiction of the Supreme Court, which mostly rules according to common law norms, over the Land Title courts, which mostly rules according to customary norms. Most land in Samoa is held according to a form of native customary title vested in the matai (the heads of village families and local elders). The removal of jurisdiction means that the Land Title courts can make decisions about land ownership against modern judicial norms including not providing written reasons for decisions and making decisions which negatively impact human rights that would normally be protected by the Constitution.

Fiame did not immediately join the upstart anti-land reform One God of Samoa Party (FAST, from their Samoan-language acronym) after her departure but took up an invitation to lead the party after the parliament dissolved late last year. When she did, she was joined by a number of other opposition parties.

In the lead up to the elections, there were so many nomination disputes before Samoa’s courts that PacLII (the database which hosts judgements) had to make a new section for them. In a number of these cases the Electoral Commission had attempted unilaterally to disqualify a number of candidates in a manner described by the Supreme Court as ultra vires and unlawful”. In others, there was confusion over changes to the matai status and monotaga (community service) requirements.

In parliamentary democracies similar in structure to Samoa (including Australia, New Zealand, and Fiji), nomination disputes of this nature are usually adjudicated after the election in a Court of Disputed Returns (usually the country’s Supreme Court acting under that name) with a by-election or countback to fill any resulting vacancy. Section 9 of the Samoan Electoral Act provides that challenges to nominations can be done only in Court in the three days after the close of nominations. Part 14 of the same law sets out a procedure similar to that described for other countries.

Throughout the campaign, Prime Minister Tuilaepa sought to exaggerate his chances of winning insisting that he would win forty-five seats, saying that those who would prevent this landslide win will have “interfered in God’s work”. He demonised members of the FAST party as “better liars than Satan” whose “brainwashing” campaign roadshows (comparable to a town hall) he suggested were the result of “foreign” ideas and should, accordingly, be harassed by HRPP supporters. On the last sitting day of parliament, Tuilaepa attempted to order an inquiry into whether Fiame and four of her FAST party colleagues, who had been holding their rallies instead of attending the last few sittings of parliament, had committed treason. It is not usually the place of a parliamentary inquiry to determine whether someone has committed a criminal offence. Even if Tuilaepa’s motion had succeeded, the inquiry would have lapsed at the end of the day when Parliament was dissolved.

Elders in a number villages ordered the removal of FAST campaign signs to “keep the peace”, in some cases on the explicit grounds that the village council “supports the [HRPP Government] and they do not welcome FAST or any other party into [their villages]”. Matai endorsements of candidates and parties can be a powerful influence, with those challenging nominations in court often being banished from their villages.

The results and the downsides of first-past-the-post

Back to top

On election night, although the HRPP won 55% of the popular vote, both it and the FAST party won twenty-five seats, one short of a majority, with an independent taking the final seat. Samoa uses a first-past-the-post electoral system (FPTP) in which the individual candidate in each constituency with the most votes, whether or not the majority of the constituency’s voters would support that candidate.

The table below shows the national results of the election, how a system of proportional representation (PR) could have distributed the seats, and how the seats were actually distributed under the first-past-the-post system.

PartyPopular vote (%)Estimated PR seatsActual FPTP seats (diff. from PR)
HRPP55.38%2825 (-3)
FAST36.57%1925 (+6)
Independents*4.53%21 (-1)
Other3.52%20 (-2)
Total100.00%5151
The results of the 2021 Samoan general election with an estimation of seats based on proportional representation

As you can see, FAST was significantly overrepresented and the HRRP was narrowly denied the majority of the seats. If we take a closer look at the electoral system and how candidates were nominated, we will get a better understanding of what happened here.

The fifty-one Fono constituencies are required by law to consist of specific villages and localities, which do not appear to have been selected according to population (unlike the UK’s system). One result of this choice is that the constituencies are of unequal size and, as a result, HRPP voters in more populated constituencies votes were not represented in the number of seats. Indeed, the Electoral Commission’s results show that some constituencies had almost four thousand voters and others only had about one-and-a-half thousand.

Normally, when political scientists talk about the downsides first-past-the-post, they point the significant overrepresentation of the winning party and the under-representation of other parties especially where the first-place party wins a majority of the vote. An extreme example of this phenomena occurred in the Canadian province of New Brunswick in 1987 where the Liberal party won all 58 seats with only 60% of the vote. The opposition NDP and Conservative parties did not win a single seat despite having almost 40% of the vote between them.

Unusually, with the recent Samoan election we are talking about the under-representation of the winning party and the over-representation of the opposition parties. This is not the result of electoral fraud but is a consequence of the FPTP system itself. What is different about the Samoan election from the 1987 New Brunswick election is that the HRPP ran multiple candidates in almost every constituency.

By contrast, FAST and their allied parties pledged to run only one candidate in each constituency. This was vitally important as in a first-past-the-post system, a party running multiple candidates in a seat can lead to the vote splitting between supporters of the respective party to the point where a candidate that the majority of the voters do not support can win.

Take the constituency of Faleata 1 for example, where the results looked like this:

PartyCandidateValid Votes
HRPPLEAPAI Richard S. Brown1013 (27.40%)
HRPPLEPOU Petelo II76 (2.06%)
HRPPMANULELEUA Ioane K. Manuleleua220 (5.96%)
FASTMANULELEUA Paletasala T. Tovale ✅1207 (32.65%)
HRPPSALAUSA John Ah Ching1181 (31.94%)
Total3697 (100.00%)
Results in the Faleata 1 constituency via Samoa Electoral Commission on Twitter

In this constituency, the HRPP won over two-thirds of the vote but their vote was split between four different candidates. This meant that no HRPP individual candidate had more votes than FAST’s singular canidiate, handing FAST the seat under the first-past-the-post electoral system. Had the HRPP’s members in Faleata 1 compromised and endorsed one candidate, they would have won this seat in a landslide.

There are a variety of solutions to this problem. The first would be to limit the number of candidates a party can endorse in each constituency. This can be enacted either through each political party’s rules or, as is the case in Australia, a law voiding multiple nominations by a single party in a constituency. If parties still wish to nominate multiple candidates, a proportional representation system, such as single transferable vote, could be used. This would prevent a recurrence of this specific outcome. Another solution would be to adopt a preferential voting system, such as instant runoff voting (IRV) so that supporters of losing candidates can redirect their vote elsewhere.

The aftermath and the myth of ‘dominant-party democracy’

Back to top

In a consolidated and stable parliamentary democracy, even with results as distorted by the choice of voting system as this, there would not ordinarily be a problem with a hung parliament. The major parties would simply negotiate with the crossbench (in this case, the single independent), come to some kind of confidence-and-supply deal, and form a minority or coalition government.

This is what the HRPP themselves had done in 1996, when they reached a supply deal with ten independents, and in 2001, when most of the twelve independents who won seats joined the party, in both cases giving the HRPP government supermajority control over the Fono.

However, Samoa is not a consolidated democracy. It is even not a ‘dominant-party democracy’, as supporters of normative measures of democracy would like you to believe, because there was absolutely no pretence that the ruling HRPP government were planning to step aside if they lost. Samoa is a dictatorship and the events described in this section will only demonstrate why PACL were right to classify normatively “free” democracies as such. On page 12 of Classifying political regimes, PACL offer this knowledge:

Even when incumbents hold elections only because they expect to win, they sometimes make mistakes. They hold an election and lose. Then they have to decide whether to accept the popular verdict or override it. They can revert to a post-election fraud [as in the case of Mexico’s PRI] … Or they can publish the voting results and still not allow the opposition to assume office.

M E Alvarez et. al. 1996. “Classifying political regimes” in Studies in Comparative International Development no. 31: page 12

In the same article, PACL go on to write about Botswana, a similar case of a type-II dictatorship, to justify why their observational approach to classifying democracy, rather than other measures’ normative approach, is more applicable:

Elections may be held in Botswana only because the ruling party is sure to win, but how are we to know what would happen if they expected to lose or in fact lost?

Ibid.

As discussed in a previous section, Tuilaepa gave no sign that he was going to leave office at the end of April and frequently tried to frustrate the opposition’s electoral efforts. Can it really be said that he expected to lose? No. And the expectation that he would not was confirmed by the aftermath in which there was deadlock and a pressing constitutional question that Tuilaepa would cling to in an attempt to overturn the election.

The battle for the premiership

Back to top

The crisis began when the Electoral Commissioner appointed a member of the HRPP to as an additional MP citing Samoa’s 10-percent minimum parliamentary gender quota. The gender quota amendment to the constitution requires that the Electoral Commissioner appoint the female candidate who came closest to winning as an “additional member” where less than 10% of the seats are filled by women.

The Electoral Commissioner’s determination would have given the HRPP a slim majority of seats, but the independent MP joined FAST, and so the parliament became deadlocked again. A similar appointment occurred without incident in 2016, where the HRPP had already won a majority and, consequently, the additional member did not fundamentally change the result.

The Supreme Court, whose judges were all appointed on Tuilaepa’s advice, ruled the 2021 appointment to be unlawful. The Court’s reasons are not yet available on PacLII as I’m writing this, however I suspect it has to do with rounding.[note]

Five women were elected on polling day, filling 9.8% of the seats. Tuilaepa insists that rounding is unacceptable, saying that the court decision shows that “the Supreme Court cannot divide properly”. From Tuilaepa’s statement, it appears that the judges were also divided (no pun intended) on the question of whether the constitutional quota can be met by rounding the decimal up. Reporting from the courtroom suggests that the Electoral Commission struggled identify the legal process for the appointment of the additional female member.

It is not unusual, in the manner Tuilaepa suggests in the quote below, for courts to be divided on a constitutional question and Tuilaepa’s own words suggest that the decision to drop the additional member was a 2-to-1 majority decision.

“Firstly the Judges’ conclusions were inconsistent. [Justice] Vui said there should be six, while [Justice] Lesatele and [Justice] Niava said there should be five,” 

Tuilaepa as in S Wilson. 2021. ‘Court “severely undermines” constitution: Tuilaepa’ in Samoa Observer.

Tuilaepa also advised the ceremonial head of state, the O le Ao le Malo, to suspend parliament and tried to call a snap election. The Supreme Court again (in another yet to be published decision) quashed this order as unconstitutional because the suspension would have meant that the Fono elected last month would not have met within the constitutionally required 45 days after an election.

Parliament was supposed to meet a few days ago, however, The Speaker – himself a member of the HRPP – ignored the Court order (and, by extension, the Head of State’s initial order to summon the Fono) by locking up the parliament building. The Head of State was nowhere to be seen, having returned to his home village.

FAST responded to the Speaker’s closure of the Fono by holding its own swearing-in ceremony under a marquee outside Parliament, with Fiame justifying the extraordinary ceremony under the doctrine of necessity (a view supported by legal scholars). The Speaker’s actions have meant that the 45-day requirement has lapsed, a fact which Tuilaepa wishes to use as justification for a snap election.

What now for Samoa?

Back to top

In a consolidated democracy, the matter should have been settled, at least for the time being, with the Supreme Court decision late last week. The legal question over the quota has been resolved, the Fono has been summoned, and a majority government has been declared duly elected. But Samoa is not a democracy. It is a type II dictatorship, and if it continues down this path, its dictatorship status will be indisputable.

Tuilaepa and his supporters have made concerning interventions including: floating plans to ban Facebook during the second campaign to prevent people criticising the government with “unfounded allegations”, Tuilaepa announcing that he was “appointed by God”, accusing the judicature (without evidence) of engaging in a “conspiracy” with the opposition, announced a prohibition of new candidates nominating in the ultimately cancelled second election, and once again accused Fiame and FAST of treason.

A former police officer initiated a private prosecution against leading members of the FAST party for bribery in an attempt to stop them taking office. The Attorney-General has acted on the bizarre “conspiracy” accusation by applying to have the entire Supreme Court recuse itself from any further election cases.

It will all come down to the head of state, the O le Ae le Malo, who has been remarkably absent from the chaos, who will ultimately have to swear in a government. He was not present at the swearing-in on Monday as he remains in protective custody in his home village due to alleged threats. He has not been heard from since his order for a snap election was set aside.

The police, who constitute the closest thing to an army that Samoa has, have publicly committed themselves to remain independent. The police are not involved in the private bribery prosecutions. In other jurisdictions, prosecutions of this nature rarely succeed (see Sankey v Whitlam for an Australian example of a private prosecution of a politician for alleged corruption).

The Supreme Court has, thus far, not accepted the various legal arguments brought by the HRPP, the Electoral Commission, and other government parties to election disputes. Various governments in the Pacific, starting with the government of Micronesia, have already recognised Fiame as the legitimate Prime Minister and called for peace. The Court will, in its appellate jurisdiction, hear a final appeal on the gender quota issue on Monday.

At least one HRPP candidate has indicated, should there be a snap election, he will join FAST if elected, suggesting that Tuileapa is facing a rebuke from within his own party.

Conclusions

Back to top

As for the measures of democracy, the crisis in Samoa, in which the country’s first transition of government at an election has been compromised, demonstrates that the observational approach of DD is better than more normative approaches to democracy like those of Freedom House.

The Samoan election was complicated by legal confusion over a gender quota, the use constituencies of unequal sizes, and the nomination of multiple candidates by the Human Rights Protection Party in most constituencies. As a result, Samoan election was a unique case of first-past-the-post resulting in the party winning the majority of the vote being underrepresented.

Although Samoa has various civil liberties, a person informed on the background to the 2021 election should not have expected Tuilaepa to respect democratic processes given that he was singing about how he would win in a landslide and had constantly denounced the opposition as traitors. As such the idea that Samoa was a ‘dominant-party democracy’ as opposed to a dictatorship is ludicrous. The transition to democracy will only be complete when there is a peaceful change of government. This has not yet occurred.

There are promising signs that Samoa’s independent judiciary will break the impasse, but it is impossible to predict at this point in time.


Notes

There are certainly other criticisms of DD that are worth discussion in this context, such as how it treats non-partisan democracies (like Samoa was before 1982), but they are beyond the scope of this post. ↩︎

An alternative to the ’rounding’ explanation for the gender quota court case above, from an article from RNZ, suggests that a provisional clause stating the number ‘5’ as the minimum number of female was still in effect. ↩︎


About the author

Joseph Mann Avatar