The population of New Caledonia is deeply split on whether to become independent from France, a problem the upcoming referendum is unlikely to solve, writes Pierre-Christophe Pantz.
On 4 November 2018, New Caledonia held the first of three referendums on self-determination. Defying opinion polls which predicted a very heavy defeat for the independence camp, only 56.7 per cent of voters voted against independence, a majority of 18,535 votes.
This surprisingly strong result increased hope amongst independence voters that they might go one better in the second referendum, to be held on 4 October 2020. However, this pro-independence momentum was weakened six months later in provincial elections held on 12 May 2019, where the pro-independence vote dropped slightly.
Close analysis of the pro-independence vote for about 30 years confirms that successive provincial elections are an important barometer for the future prospects of self-determination. This would indicate that the pro-independence vote is relatively unlikely to obtain a majority in the two upcoming referendums. This reflects the entrenched nature of the pro-and anti-independence votes, neither side having significantly strengthened their bases in 30 years.
Additionally, voter turnout is set to be a major sticking point for this referendum. In 2018 turnout reached a record-high 81 per cent. Over the previous 30 years, it had never exceeded 77 per cent, and had decreased continuously during all provincial elections over the past 15 years. Notably, in some pro-independence strongholds the 20-year deferral of the referendum process through the Noumea Accord led to a disengagement of the pro-independence movement.
The Kanak vote is also likely to play a major role in 2020, as it did in 2018. The strong correlation between the indigenous Kanak population and the pro-independence vote, constant since pro-independence parties emerged in the 1970s, continued for the referendum and the provincial elections. This is likely to persist over the course of the next two referendums, and as a consequence, a shift in the political balance is unlikely.
This correlation is visible when comparing census data that shows the geographic distribution of Kanaks with data from polling stations, indicating the spread of pro-independence votes by polling station. For example, the polling stations with the highest pro-independence results form a cone-shaped area in the northwest of Noumea. This cone coincides with an underprivileged area where a high proportion of Kanak people reside.
In addition to geographic and ethnic voter patterns, another predictor of voter preference is the village-tribe divide. This especially notable in rural areas, between voters in village centres, who are mostly non-Kanak, and voters outside of village centres, who are mostly from Kanak tribes.
The clearest case of this divide is in the far north village of Ouegoa, whose population is 72 per cent Kanak. 30 per cent of the village’s population live in the centre, with three out of four voters voting ‘no’ to independence. 70 per cent of the town’s population live outside the centre, in the area covered by three polling stations. Only Bondé, Paimboas and Tiari tribe members are enrolled here; and only 42 in 1,148 voters – 3.7 per cent – voted against independence.
Shifting political spheres of influence will also have a noticeable effect. Provincial elections on 12 May 2019 brought in the narrowest political gap – 26 pro-independence to 25 anti-independence members of Congress (out of a total of 54) – since New Caledonia was divided into provinces. However, three elected members of Congress have chosen to remain neutral. The pro-independence group also strengthened its base in Congress as it now holds the key positions of president and vice president, and controls the permanent committee and internal committees.
However, despite this newfound strength in Congress, the distribution of the pro-independence vote remains unchanged. What is notable, however, is the fact that the pro-independence parties secured 48.1 per cent of the congressional seats with only 42.4 per cent of the votes. This over-representation of pro-independence parties at the institutional level has two major causes, a distortion in representation and ‘the five percent rule’.
When New Caledonia was divided into three provinces, a distorted representation was implemented to the benefit of the Northern and Loyalty Islands provinces. Both provinces have proportionally more seats in Congress than the Southern Province. A strictly proportional representation of the population would give Southern Province 40, Northern Province 10, and Loyalty Islands four seats. Based on electorate only, the seat numbers would be 34, 13 and seven. For comparison, the actual number of seats is 32, 15 and seven. This distortion favours the pro-independence representation in Congress.
In addition, political parties have to follow the ‘five per cent rule’ in order to be represented in Congress. Since 1999, parties need to obtain the votes of at least five per cent of registered voters in order to win a seat in Congress. Votes in favour of parties that fail to meet this requirement may be deemed ‘useless’, as they don’t materialise into seats. This makes the number of seats that will be won by either side difficult to anticipate.
The main consequence of the ‘five per cent rule’ is stasis in the political landscape, which benefits historic parties, but hinders smaller or newer ones. The votes garnered by parties that fail to meet the required threshold are not taken into account in the allocation of seats and are discounted. Since the 2004 election, the pro-independence parties have greatly reduced their number of ‘useless votes’, mostly in Southern Province, where they have opted for a unitary approach so as to maximise their representation.
The votes lost by the anti-independence parties, however, have not only considerably increased in Southern Province, but also in Loyalty Islands Province, where they are currently not represented in the provincial assembly. Consequently, the difference in congressional representation does not indicate a shift in the political balance in New Caledonia.
Two certainties emerge from an in-depth look at voting patterns from November 2018 to May 2019. First, voting patterns have frozen. There are very few towns that switch from pro-independence to anti-independence come election time. Indeed, Kanak and pro-independence strongholds were built based on this stability. This might explain why election turnout rates are decreasing.
Second, 30 years of political stability and balance have failed to change the ethnic and geographical nature of the vote. The Kanak people continue to overwhelmingly back independence, while non-Kanak people keep voting to remain in the French Republic. On both sides, the challenge of reaching beyond ethnic boundaries has not been met.
The map of abstention from the referendum, which correlates with the pro-independence vote, may be a sign that the gap is narrowing for the coming referendums. This would require a maximum turnout on the pro-independence side and the hope of a diminished commitment on the anti-independence side. Only then would pro-independence parties be able to break through at the referendum.
However, a narrow victory for the pro-independence parties would not be the solution to the New Caledonian conundrum. Together with those who are not allowed to vote – 36,000 people for the 2018 referendum and 41,000 for 2019 provincial elections – those against independence will make up more than half of the population of New Caledonia. It would be the same if independence fails at the second and third referendums – a significant part of the population will still support leaving the French Republic.
The stasis of the New Caledonian political landscape leads to a double dead end. It seems that, in all likelihood, only a shift in the political paradigm, breaking with a polarised opposition, would offer a realistic and satisfying way out for the greatest number of people.
This article is part of the ‘In Brief’ series and is based on Department of Pacific Affairs (DPA) Working Paper 2020.05, which is available here.