One Man One Vote One Value: Part I, Pegs and Holes
Trying to Put Round Pegs into Square Holes
by Crosbie Walsh
(Readers using an iPad or smartphone will not be able to see the maps and tables in this article.)
Use CTRL and + or - to increase or decrease the size of maps, tables -- and text!
All previous elections in Fiji have been based, to a greater or lesser extent, on its 14 provinces with a reserved parliamentary seat for the island of Rotuma and Rotumans living in other parts of Fiji (Map 1.1). In this article I will argue that all previous elections have been intrinsically unfair because of this reliance on provinces, and that the only way to achieve one man one vote one value is to dispense with the provinces as the geographic unit around which the electorates are drawn.
The columns in Table 1 show the very unequal population sizes of the provinces and the equally unequal distribution of ethnic groups within each province. For example, Ba has 27.7% of Fiji's population and 40.2 % of Indo-Fijians, while Lau has only 1.3% of Fiji's population and only 2.2% of its indigenous Fijian population
By contrast, the rows in Table 2 show Indo-Fijians comprised only slightly more one half (54.5%) of Ba's population while 98.l7% of Lau was indigenous Fijian.
These geographic differences account for much of the difficulty in coming up with a fair electoral system based on ethnicity, if the system is based on provinces.
In recent elections 46 seats have been allocated to Communal electorates and 25 to Open electorates in the 71-member parliament. In this article we will look at the Communal electorates, and in the next article at the Open electorates as a better way of achieving voter equality. The maps are from my book Fiji: An Encyclopaedic Atlas, and the election results shown are for 2001.The tables are based on 2006 election and Bureau of Statistics data. MV in the map below stands for Matanitu Vanua (CAMV), the extremist iTaukei nationalist party whose members joined Laisenia Qarase's SDL party in the 2006 elections.
The 17 Fijian Communal electorates were based on single provinces (Map 16.4). Each province received one seat with the exception of Ba, Cakaudrove and Tailevu which were each given two seats in recognition of their larger populations. This adjustment still resulted in some people's votes being worth far more than others. The 19,000 voters registered in Nadroga/Navosa, the 15,400 in Ba West and the 11,600 in Cakaudrove West, for example, each elected only one MP as did the the 6,000-odd voters in Bua, Kadavu and Lau, the 4,500 in Serua and the 3,000 in Namosi. A vote in Navua was worth six-times more than a vote in Nadroga/Navosa.
The six Fijian Urban electorates (Map 16.5). were divided equally between urban places in the North-West (Ba to Nadi), the North-East (Tavua to Nausori and Vanua Levu), the South-West (Sigatoka to Lami), with three seats allocated to Greater Suva urban area. The number of registered voters was reasonably similar, ranging from 15- to 18,000, much larger than the rural Communal electorates that averaged 9,500 voters, but urban Fijians were considerably under-represented compared with their rural cousins. Of the 23 seats allocated to Fijians, 62.5% went to the rural and 37.5% to the urban seats, but only 55.5% of Fijians lived in rural areas compared with 44.5% in urban areas (2007 Census).
This rural-urban discrepancy favoured the generally less educated, those less involved in the modern money economy, those with fewer opportunities to meet and mix with people of other races, and those more easily influenced by their local chiefs and church ministers. Local, traditional and racial loyalties would displace all others.
The 19 Indo-Fijian Communal electorates (Map 16.6) displayed similar proportionality problems to the Fijian electorates. The average number of registered voters was similar (10,800 to the Fijian 11,200) but the need to combine and divide provinces to obtain approximately equal numbers in each electorate, compounded by the very uneven distribution of the Indo-Fijian population, resulted in a wide range in the number of voters per electorate. The extremes were VitiLevu East/Maritime, that encompased parts of Ra and Tailevu provinces and the whole of Lomaiviti and Lau, with under 8,000 voters and Laucala, in the Suva urban area, with over 18,000.
The three General Voter electorates (Map 16.6) and indeed the very concept of general voters, must raise eyebrows. The total number of registered voters was 13,800, far less than the number of Indo-Fijian voters in the one electorate of Laucala mentioned above, but they were allocated three seats: the West-Central that included most of Viti Levu, Rotuma and Kadavu; Suva City,excluding Lami, and North-East that included part of Tailevu and all of Vanua Levu, Lomaivit and Lau. The average electorate size was 4,600. In the 2006 election they comprised only 2.9% of registered voters but occupied 6.5% of the communal seats in Parliament.
In many respects the provision for General voters is a leftover from colonial days when Fijians were outnumbered by Indo-Fijians and the British thought General voters would ally themselves with Fijians and so protect Fijian interests. In the census they are called Others, which accurately reflects their composition. They comprise Europeans, Part-Europeans, Chinese and Part-Chinese, KaiSolomoni, Kai Vanuatu and other Melanesians, Tuvaluans, Banabans, Tongans, Samoans, and other Pacific Islanders. As I have said elsewhere, they are not a race or an ethnic groups and it is doubtful they have a common purpose or can be represented by a person from only one of these groups. I doubt, for instance, that the KaiSolomoni, the Banabans on Rabi or the Tuvaluans on Kioa Island identify in any special way with the Part-European Mick Beddoes who represented them in the 2006 Parliament.
Mick says he has "a simple solution" about how to ensure parliamentary representation for ethnic minorities. He says let's assume the minorities comprise four percent of the population and merit three seats. “So we divide Fiji into three but every citizen in Fiji votes for that minority.” Having iTaukei and Indo-Fijians voting for the “minorities” is presumably his sop to multi-racialism. The fact that the minority vote would be swamped by the larger numbers would complete their disenfranchisement seems less important than the retention of the parliamentary seats.
Other Races is a a patched together et cetera, a left over. If they require separate representation in parliament, they would be better divided into Part-Europeans of whom there are about 11,000, Other Pacific Islanders (10,000) and Europeans/Chinese and Part-Chinese (9,000) with one national seat each.
The Rotuma electorate represents about 2,000 Rotumans on the island of Rotuma, some 390 kilometres north of Viti Levu, and approximately 8,000 living in other parts of Fiji. There were about 5,400 registered registered voters in 2006. On the basis population numbers, they were significantly over-represented in Parliament.
In sum, these communal seats elected 46 MPs on very unequal grounds. Rural Fijians were over-represented, with 37% of the seats but only 33% of the population; Indo-Fijians were slightly under-represented (41.3 and 42.6%); Urban Fijians were substantially under-represented (13 and 20%); and Rotumans (2.2 and 1.1%) and General voters (6.5 and 2.9%) were substantially over-represented.
It is difficult to see how the one man one vote one one value principle can be achieved if race and the communal system of elections is pursued. Far better representation will be achieved if the Open electorates are used a guide, but even then some tinkering may be necessary if "special interest" groups such as the ethnic minorities are to be considered.
Ratu Tevita Moemodonu's submission to the Constitution Commission