Cogito, ergo sum. I think, therefore I am. (René Descartes, mathematician and philosopher,1599-1650)

Thursday 1 October 2009

(B) Election Formats: Part I Narsey on Proportional Representation and the List System

In this two-part series USP economics Prof. Wadan Narsey and I come to rather different conclusions on the possible outcomes of electoral reform. Wadan's two related articles, Electoral Reform Not about Racial Justice and Does the Military Really Want Electoral Reform? were previously published by Pacific Scoop. I abridge and summarize their contents here with a link to the originals. Part II Walsh New Election Playing Fields and Outcomes will be published next week.




Electoral Reform Not about Racial Justice

Wadan looks back and forward at proportional representation
To see the full paper, click here.



Government supporters state there should be no elections until a racially fair one-person one-vote system is in place. Only then, with a proportion list system in place, will electoral racism against Indo-Fijians cease.

But an analysis of the 2006 results shows that the old system was not racially unfair, although it did have other weaknesses. It was unfair to smaller parties and was too complicated for voters. The proposed proportional voting system will not make any great difference from an ethnic point of view, although it will be much fairer to the smaller parties. Indeed, in all previous elections the proportion of parliamentary seats held in aggregate by the Fijian and Indo-Fijian parties have been roughly the same as their share of the votes.The existing electoral system has been roughly “racially fair”. The real unfairness was that none of the smaller political parties both Fijian (NAP, PANU, SVT) and the Indo-Fijian party (NFP) could obtain any seats at all in 2006 (because they failed to win more than 50% of the votes in any constituency) while the larger parties were over-represented. The SDL's 44% of the votes gave it 52% of the seats, and the FLP with 40% of the votes had 42% of the seats. The proposed proportional/list system will eliminate this unfairness. In this article Wadan looks at proportional representation; in the second article at the List system.

How will a Proportional System Work for Fiji?

The number of electorates do not matter because each party’s share of the national votes cast will determine that party’s share of seats in the Parliament. There could be five, as suggested by the NCBBF, 25 or any other number. Each voter will be given only one ballot paper with the names and symbols of all the political parties wishing to stand. They will mark just one box, with a tick, or a cross, or a slash or whatever, for their preferred party. Nothing could be simpler.

Wadan examined the effects of proportionality by using the first preference votes only in the 25 Open electorates (which had common rolls) used in the 2006 election to see what the outcome would have been had there been no ethnic Communal electorates; just the Open electorates. He found that the SDL would have had 33 seats (not 36), the FLP 30 seats (not 31) and that two parties that had no seats in 2006, the NFP and the NAP, would have had 5 and 3 seats respectively. He ignored the 17,284 votes (4.5% of valid votes) cast for Independents and could, of course, do nothing about the 63,164 (15% ) of registered voters who did not vote, or the 37,242 (8.8%) of registered voters whose votes were declared invalid. Using the same methods, I arrived at the same result. In other words, it is reasonable to assume that irrespective of whether the old Alternative Vote or the proposed Proportional method is used, the SDL would have won the 2006 election, unless the NFP and/or NAPF came to some agreement with the FLP .


Does the Military Really Want Electoral Reform?
Wadan looks at proportional representation in combination with the proposed list system,
and the fate of Independent, General and Rotuman voters. To read the full paper click here.

He argues that the NCBBF Draft Report’s recommendation for four or five large constituencies has two weaknesses: voter could be faced with a hundred names to choose from, and there would be no locally accountable parliamentarians to serve purely local needs for roads, water, sewerage, health, education, and so on. It thinks it would be much better to have a larger number of “local” constituencies to ensure that all our people have a specific local parliamentary representative to go to for their purely local needs.

How Does the List System Work?

Before the elections, all political parties will publish a “List” of all their candidates, ranked in order of importance, hopefully those with expertise to be Ministers, at the top. Successful List candidates will be chosen from the top, going down to the required number. It will be clear from the List order, how much the Party values ethnic balance, gender balance, regional balance, age balance, etc.

Each voter will be given two ballot papers: one for the political parties on which they will tick the party of their choice; the other for the local constituency. Each party's share of the total party votes will determine its share of seats in Parliament (minus the Independents).

In a parliament with 71 seats, made up of 25 common roll constituencies and 46 MPs selected from their party lists, an analysis of the 2006 election shows SDL had 46.8% of the total party vote. This would entitle it to 33 seats. If it won 14 constituency seats, it would be entitled to another 19 seats from its list candidates to make up its total of 33. FLP, with 42.2% of the party vote, would be entitled to 30 seats. If it won 11 constituency seats, it would select the top 19 candidates on its list. Neither NFP, with 6.5% of the total vote, and NAPF with 3.6%, won a constituency seat but their proportion of the total party vote would entitle these parties to 5 and 3 seats from their party lists.

The great advantage of the proportional List system is that smaller parties do not have to win anywhere, but their national votes will still be aggregated, and if they have enough votes nationally, they will get the appropriate number of seats in Parliament. You can move the boundaries anywhere you like. Fiddling or “gerrymandering” boundaries will have very little effect. It does not matter what the local results are. If by some accident, SDL had won all the 25 local seats (and FLP none), then SDL would only select 8 additional persons from their List to make the same national total of 33 in Parliament. If FLP won none locally, they would select all 30 persons from their List to make the same total of 30 in Parliament.

What of Independents and the “Generals”? If Independents win locally (say 3 seats) then the national total for sharing between the political parties is reduced by 3 seats, i.e. the political parties will share 68 seats to get their overall totals. Parties representing General voters would have needed to obtain roughly 6,700 votes to get one MP into Parliament. In 2006, there were some 13,800 General voters. If they all voted for the same person, they would get at least 2 seats in Parliament. But Rotumans may struggle to elect anyone using their own votes alone.

Who forms government? In the normal Westminster system followed by most countries (US, Australia, UK, NZ…) any party getting the support of more than a half of Parliament (in Fiji, 36) forms government and the others become the Opposition.

With the above results, both SDL and FLP could form government, depending on who the smaller parties support- smaller parties will indeed have great power in such situations. Of course, SDL and FLP could also form a Coalition and form Government! (dream on). Or Fiji could still retain the Multi-Party Government provision (which all our recent political leaders have miserably failed to use and thereby helped to ruin this country).

Readers can read my previous published articles on these questions, at this link. Photo: USP 2009.

No comments: