An Accountable Electoral System
|Father David Arms (right)|
For the 13th Attorney-General’s Conference, 2nd & 3rd December 2011
In discussing the topic An Accountable Electoral System, this paper will look at three different aspects of it. Firstly there is the need for the people as a whole to regard the adoption of a new electoral package as legitimate and desirable. Secondly, there is the need for the various electoral processes, from the registration of voters to the announcing and implementing of the electoral results, to be checkable and transparently fair. Thirdly, there is the need for the actual electoral system itself – how votes are cast, tallied and the results arrived at – to make sense, so that the electorate can feel confident that its voice has been accurately conveyed. The paper will conclude with some extra observations relevant to the theme of accountability.
1.0 Whole Enterprise Accountable to Fiji’s People
1.1 A country’s electoral system exists primarily for the people it serves, in order to provide them with the representatives they want. If an electoral system is to be regarded as “accountable” therefore, it must be one that first and foremost passes muster with them.
1.2 No doubt other countries will have an interest in the matter too. They will want to be assured that the electoral system is suitably democratic and that the whole process is free and fair. Electoral experts will also have an interest, and may look to promote even higher standards than foreign governments. However, if a system of national elections is adopted which the populace is happy with, this will usually be enough to satisfy foreign interests, both governments and experts.
1.3 When, therefore, we are talking about an accountable electoral system for Fiji, our first priority must be to ensure that Fiji’s people as a whole are happy with the overall process and with the particular system to be adopted. Choice of such a system is indeed a challenge. For it is not a matter of simply asking the people, “Do you like this electoral system, or that?” Electoral systems are complex entities. The ordinary citizen cannot be expected to have a knowledge of the various existing electoral systems and their variations so as to be able to decide from among them which is to be preferred. So how can they be expected to give reasonable input or assent to the system to be adopted?
1.4 Obviously, experts, both foreign and local, have a role to play, but the decision cannot be left to an elite of this kind alone. There needs to be ownership of any decision by the grassroots. It would seem most practicable that some sort of Assembly, broadly representative of the people, be established, which would be the body that meets, discusses and decides on this issue. Since Fiji also needs to settle on a new constitution at this time, the two processes could well be conducted together. There has been talk of setting up a Constitution Commission. Such a Commission may indeed have a role to play, but it is too limited in its composition to do the electoral job effectively. It is a wider group – a Constitutional Assembly, perhaps led by such a Commission – which would take basic responsibility for the task, even if it passed over a lot of the detailed work to a sub-committee.
1.5 In being an acceptable group in the eyes of the people, the Constitutional Assembly would provide the desired accountability. Although a referendum remains a possibility, the complications of this (e.g. would the referendum be on the whole Constitution, or on one particular electoral system only, or on choosing between systems?) may be able to be avoided – so also the expense, although the degree of consultation suggested in 1.7 below would itself, of course, be expensive. In any event, the need or otherwise of a referendum would be clearer as the Constitutional Assembly’s work progressed, so it could be left to that group to decide for or against one.
1.6 Some of the way for this Constitutional Assembly has been paved already. The group that was assembled and produced the Charter is the type of group that might do the job again. There was, of course, at the time of that group’s operation, considerable controversy about its representativeness and objectivity. As time has progressed however, and as one takes a good look at the fruit of that group’s labours (the People’s Charter and the State of the Nation Report), and at its electoral proposals, it is apparent that good directions have already been set. It is important that the good work already done be acknowledged by picking it up and carrying it forward to completion, incorporating as appropriate further ideas or modifications from a wider group.
1.7 The Constitutional Assembly therefore that I have spoken of, could in fact be a partial recall of the Charter group, with the very deliberate inclusion of further members to represent groups who were not part of the Charter process, whether through being overlooked, or having excluded themselves from it. If we are to have an accountable electoral system, it is extremely important to have the group basically responsible for it one that is acceptable to the people, responsive to it, and respected by it. Furthermore, no such Constitutional Assembly should be operating just on its own internal dynamics, but on continuous feed-in from, and feed-back to, the general populace. The Charter consultations provide some model for this, and the whole process will constitute an initial stage in voter education.
2.0 Accountable Procedures within the Overall Electoral Process
2.1 It is important that electoral officials be competent, neutral and trustworthy. Within their respective terms of reference and budgets, the Electoral Commission and Elections Office need to be independent. They must steadfastly and manifestly keep themselves free from interference, whether by local or foreign interests, political or otherwise. It is the people’s interests as a whole that they must be concerned about and held accountable for. Good procedures, then, for the selection of suitable people (including for short-term work) need to be in place. Good personnel, appropriately trained, are vital. It is likewise important to check out the large number of steps within an election, each of which needs to be done professionally and accountably. Below are some of the more important ones:
2.2 Clear and fair laws will need to be passed about who will have the right to vote and who the right to stand. These should not be too restrictive, as the whole idea of elections is to gauge the will of the people – the whole people as far as possible. Even people with extreme viewpoints should usually be heard, and as regards those who are standing, surely the people can be trusted not to vote for unworthy candidates. Nevertheless, some regulations are needed. Rules for the operation of parties are also necessary. Again, these should not be too restrictive, as the forming of political parties and voting on them form a basic part of democratic freedom. Up to a point, it is better to have objectionable parties rejected by the people (through voting) rather than ejected previously by law from the electoral process. They should be required, nevertheless, to espouse democratic values, including fairness and transparency.
2.3 Related to this, there will need to be clear and fair laws governing campaign funding. Such laws need to be extremely well thought through so as not to leave loopholes. This is an area where strictness should be applied. Vote-buying in any form must be outlawed. So must the massive use of advertising and propaganda, which can so much advantage a ‘rich’ party over a ‘poor’ one. Abuse in the use of funds during elections undermines the very concept of democracy, because it can end up that results are determined more by the availability and use of funds than by the genuine consideration of the issues by voters and their expressing in true freedom their views on them by their vote.
2.4 It is important that all people with the right to vote are properly registered. (It is proposed at the moment that registration be compulsory, but voting itself be optional.) For the last four years, the Elections Office has been recommending and planning for the adoption of a system of Electronic Voter Registration (EVR). It is important to note that this is not the same as Electronic Voting, whereby voters record their votes on voting day by electronic means. This is not being advocated in Fiji, at least as yet – (see also 2.8 below). EVR is done, of course, well before voting day. The Elections Office has already researched 1,150 locations to be used as registration centres. The electronic recording of biometric data on each voter, including photo, fingerprint, and perhaps signature, makes it very difficult for voters to impersonate other voters, vote twice, or vote in the wrong polling station. Experience with EVR elsewhere confirms that it adds a lot to the accountability of the whole process, and also provides far less frustration to voters (some of whom used to find that their names were not on the roll, or were placed in the wrong area, or were mis-spelt or garbled, etc.). Each voter receives a registration card, and presentation of this card on voting day makes for easy and accurate issuance of the voting paper.
2.5 It is recommended here that, insofar as possible, voters should all vote on just one day - see 2.7 below. There do have to be exceptions, however. Some people (not only those overseas) will, through circumstances, be a long way from where they should cast their ballot on polling day. In the past, there have been facilities for postal ballots. There is a genuine security problem with this sort of voting, however. It is recommended here that postal voting be abolished and that a form of Advance Voting be adopted instead, to be used only by those who are confident they cannot go to their polling station on polling day. This might be the case not only in individual cases, but also for whole communities, e.g. those living in remote locations. The Elections Office would have certain venues designated for such Advance Voting, and would ensure that very tight security is kept on the ballot boxes. The Elections Office would also have some portable Advance Voting facilities so that they could record the votes of people in hospitals and rest–homes, or in the remote locations mentioned above, rather than expect everyone to come to pre-designated Advance Voting centres. (Similar provisions for special cases should also, of course, be made when doing the EVR spoken of in 2.4 above.)
2.6 There has been a problem in the past accounting for unused ballot papers. While there will indeed need to be strict protocols in place with regard to the handling of ballot papers (including accounting for unused ones), the Elections Office is at this point recommending that all voters have one designated polling place where they will go to vote. These polling places will be at the same locations as the 1,150 registration centre venues already researched (see 2.4). Since the maximum number of voters to vote at each venue will be known from the data compiled at the time of registration, it will be possible to deliver to each venue a quantity of ballot papers reasonably close to the total number of voters who can possibly vote there. There should not, therefore, be a large surplus of ballot papers. It will probably be advisable, for the benefit of voters a long way away from their proper polling place, to have a few central venues designated where people can go to vote instead of having to go all the way home for polling day (even though it might have been better for them to foresee this eventuality and to have done advance voting). If this facility is granted, it should not much increase the number of surplus ballot papers.
2.7 As stated above, the strong proposal at present is that voting and counting should all be done in just one day. If there is anything which caused concern in previous elections, it was the fact that, since voting took place over a period of eight days, security of the ballot boxes was a major issue. Even if nothing sinister took place in those eight days, with ballot boxes being held at and transported to various places round the country, it is well-night impossible to convince people of this. The system was intrinsically insecure. In spite of the difficulties of transport in Fiji, many large countries with similar problems but with much larger populations manage to conduct their elections in just one day. The Elections Office is convinced it can do it. There would need to be more teams trained to handle both polling and counting, but with the proposed new voting system, these teams would be smaller than hitherto. It seems clear that with 1,150 polling places and the expected number of registered voters for each not being above 600, the job can be done.
2.8 If polling stations open at 7am and close at 7pm, counting (under the proposed system) can begin immediately at 7pm and will normally be terminated within a couple of hours. Party agents will therefore be able to see the empty ballot box at the beginning of voting, watch the process of voting throughout the whole day, see the votes poured out of the ballot box onto a table, see the count as it is made, and witness the communication of the results to the central counting location. With vigilant party agents and voting observers in place, it will be very hard for any improper activity to take place. Using the processes described here makes electronic voting not just unnecessary, but also undesirable. A manual procedure which people can witness, provides much more confidence than an electronic one where they may fear some expert manipulation or technological malfunction. For vexed cases in particular, it is much better for the officials, including the judiciary, to be able to witness a manual recount if needed.
2.9 As just mentioned, the votes cast at each polling place will be counted at that polling place. Clearly, the officials at the polling stations (now also counting stations) will need to be well-trained in both procedures, but since the basic counting is very easy, this should not present any real problem. Since the electoral system being proposed is a proportional representation (PR) one, the results at each polling place (the total number of votes held by each party and by each candidate of each party) will need to be forwarded to the constituency headquarters (basically the divisional headquarters – see 3.8). It is clear that there will need to be a very secure system of communication between the polling places and their respective headquarters (whether phone, fax, email, text-message or some other means be used). The identities of the sender and receiver must be properly checked. It would probably be best that each set of results coming in be confirmed.
2.10 At the constituency headquarters, all the results will be collected and totalled. Since all the polling places will be forwarding their results within a two-hour period (say, 9pm to 11pm), there will need to be efficient means in place for receiving and processing the polling stations’ results. The actual entering and totalling of results could take a few hours. Tallying all the votes of the constituency and computing the proportions is a more sophisticated procedure requiring more expertise. This too, however, should not be a problem, as the people required to perform these operations will be few in number and based at the constituency headquarters. These procedures too are not overly complex and are such that any reasonably educated citizen should be able to follow them. What the officials do first is work out the proportion of seats to be allotted to each party in that constituency. This is done by adding together the total votes of all the members on each party’s list to provide a total for each party. From those totals, the correct proportion of seats to be given to each party is worked out using the Modified St Lagu method. Those candidates of each party who have most votes for that party are declared elected to fill the number of seats allocated to that party (see also 3.3).
2.11 As has been the case in the past, there need to be forms of legal redress in the case of controversy over results. Fair and efficient procedures need to be drawn up for the investigation of complaints, not just about the results, but about any aspect of the total electoral process which may have been perceived as not functioning properly.
2.12 However, quite separately from that, it would seem to be a good idea that the Electoral Commission set up a neutral group (of which some or all of them could be members) to examine the complete electoral process after each election and make recommendations for the improvement of any parts of the system. Perhaps this group, having consulted not only with Parliament but very widely, should be empowered to actually make such changes no later than two years out from the next due election. It would seem best to give this power to such a group, making sure its membership was well put together (e.g. a (retired) judge, electoral experts, those with electoral experience, and neutral citizens of wisdom and good repute). It would be best not to involve politicians as such, or people who are highly politicized (though both should certainly be consulted), since there would be a danger of decisions being made to benefit particular parties, whereas the effort should be to improve the system as such, attempting to be neutral and totally fair in making changes. It is for this reason too I have not suggested Parliament as the body to make the changes. Because the members of Parliament get elected by the system, there would be the danger they might try to change it to improve their personal chances, or those of their party, for the future.
3.0 The Electoral System itself
3.1 During the Charter process, considerable time and effort were spent discussing the various electoral systems that might be used in Fiji. Two of the main conclusions reached were: 1) to do away with communal seats, having all seats as open; 2) to use a system of proportional representation (PR). The application of these two conclusions would result in a system of ‘one person, one vote, one value’. I will mention two other conclusions arrived at during the Charter discussions; 1) to reduce the minimum voting age from 21 to 18: and 2) to reduce the size of Parliament by between 15% and 25% (that is, to reduce it from 71 members to between 54 and 60). In discussing constituencies and proportions below, I am presuming on the lower figure of 54 seats for Parliament. This was the figure up to 1987. Re-adopting this figure would clearly cut cost and would still provide ample representation for the people – about 10,000 voters per representative, which is very good by world standards.
3.2 There are three main types of PR system: 1) the Single Transferable Vote (STV) system, such as used in Ireland and Malta; 2) the Mixed Member Proportional System (MMP), such as used in New Zealand, Germany, and some other countries, and 3) a List system (List), such as used by Sweden, Argentina, and indeed the vast majority of countries employing a PR system.
3.3 There are in fact two main types of List system: “Closed” and “Open”. Some countries actually have a List system including elements of both types, which is therefore usually called a “Partially Open” system. A Closed List system is one where voters are presented with a List of candidates by each party. The voter votes by ticking the name of his or her preferred party. If when the votes are counted a particular party becomes entitled to, say, 12 seats, but there are 20 candidates on the party’s list, those 12 are declared elected who are at the top of the list (which, of course, has been ordered by the party). By contrast, in an Open List system, the candidates’ names are presented on the list in any order (or in alphabetical order) and voters vote by ticking the name of one of the candidates of their preferred party (not the party itself). If, as above, the party becomes entitled to 12 seats, the candidates to fill those seats will be determined according to which of them got most individual votes. In a Partially Open system, the list is ordered as in the Closed List system, but voters vote for individual candidates as in the Open List system. However, candidates must get above a prescribed minimum to be elected according to the voters’ choice, otherwise the party’s order will be followed. In practice this system often works out little different from the Closed List, and is certainly much less democratic than the fully Open List.
3.4 In the Charter discussions, a preference was expressed for the Open List system (although the Closed List was not entirely ruled out, nor, for that matter, was the MMP system; the STV system was, however, excluded as being too complicated). The choice of an Open List system derived from a genuine desire to grant more power to the people - to really improve our democracy in this direction. Under any circumstances, political parties already have a lot of power in elections. They choose the candidates and campaign as they see fit. It seems fully proper that the people should be the ones to have the final say as to which candidates from the list should be elected. Many parties too recognize it as more beneficial to party life. Candidates get elected by the people and not by currying favour with the party bosses so as to secure a place at the top of the list. All candidates on an Open List have a chance and will give their best, whereas in a Closed List, those at the bottom have little chance of success and are hardly likely to work too hard. The Open List provides maximum accountability because voters can see from one election to the next how the various candidates have performed. If a candidate has paid little attention to his electorate but the party still puts him on their list, voters can very easily shift their vote to another candidate on that list – they do not have to change party to show their disaffection for that candidate.
3.5 A sample ballot paper for the Open List system in a constituency having 10 seats is provided as an Appendix to this paper. The sample has one line of simple instructions in English only, but in reality the instructions would be given in Fiji’s other major languages too. In the sample, the parties’ names are given as colours (the Red party, the Blue party, etc). The candidates’ names are presented in random order on each party’s list. The voter votes by placing just one tick in the box to the right of the name of the candidate preferred by that voter. That is surely a very simple operation – and, of course, there would be assistance available at every polling station for anybody experiencing difficulties. Clearly for those constituencies where 20 or so seats are to be contested (see below), the ballot paper would need to be larger, but the voter’s task is still just as simple, and large ballot papers are used in many countries without problem. Though I prefer this explicit ticking of the candidate’s name, another method of voting is available if large ballot papers are not favoured. By this method (used in Finland), the parties’ lists are all displayed in each polling booth, each candidate having a different three-digit number (randomly selected) printed next to his name on the list. (This number has, of course, also been publicized during the campaign period). Each voter has a small ballot paper on which he writes his preferred candidate’s number (say, 712). Again, very simple, even if not quite so explicit.
3.6 How big would the party lists be? This depends on the number of seats in the constituency. A party’s list cannot have more in it that there are seats being contested, but it can have less. It’s up to the party. There’s little point in a small party running 20 candidates in a 20-seat constituency if it can only expect to win one or two seats. Even a big party may decide to run less than 20 if it thinks it can win, say, a maximum of 15. Of course, the less candidates a party runs may mean that it is not really appealing to all sectors of the community, and it may lose votes thereby. On the other hand, if the number of candidates on a party’s list is close to the number of candidates that get elected, the party will have got those candidates elected whom it particularly wanted to get elected, and will not perhaps have unexpected ‘outsiders’ elected by the voters. These are matters, of course, for party decision.
3.7 The Open List system as proposed in the Charter has large constituencies. In order to have proportional representation, it is of course essential to have multi-member constituencies (– if there is only one seat in the constituency, one cannot divide that seat into proportions!). The question is: how big or small should they be? The larger the constituencies are, the more proportional the division will be. Following this principle, some countries (e.g. Holland, Israel, and Paraguay) treat the whole country as just one constituency. But that is usually regarded as a bit too unwieldy, so smaller constituencies are generally preferred. The loss of proportionality becomes clear when we look at some examples. Let’s say a constituency has 100 seats, then if the votes are divided 67%/33% between just two parties (Party A and Party B) then Party A will receive 67 seats, Party B 33 – the proportions are exact. But if the constituency has only 20 seats, Party A will receive 13 seats, and Party B 7 – Party B thereby getting a little more than it is strictly entitled to. If the constituency has only 10 seats, then it is Party A that receives a bit more than it is entitled to – 7 seats versus 3. If on the other hand the constituency has only 5 seats, Party A will receive 3 seats and Party B 2 – Party B thereby doing a good deal better than is strictly warranted.
3.8 At the time of the Charter it was suggested that Fiji have just 3, 4 or 5 constituencies. If there were just four, they would correspond exactly with the four Divisions: Northern, Western, Central, and Eastern. One big advantage with this is that there would be no need of a Constituency Boundaries Commission, as the borders of these Divisions are already established. In addition, all processing of votes (totalling of votes, working out of percentages, and declarations of election) would be done by the respective Commissioners, as has been the case in the past. Such simplicity is very attractive. Unfortunately, this does lead to rather severe difference in constituency size (see 3.9 below), which in turn can lead to some unfairness of representation whereby smaller parties have some chance of representation in the large constituencies but virtually no hope at all in the smaller ones. The use of five constituencies, however, would enable the Constituency Boundaries Commission (which would come into play after all, unless the Electoral Commission took this rather easy job on) to even out the numbers in each constituency so that each of them would have between 8 and 14 candidates each, ensuring greater fairness. The Northern constituency would stay the same, the Western and Central would be divided in half, and the Eastern would be combined with one or more of the others, (giving five constituencies in total). The counting would still be done at Divisional level, making due adjustments.
3.9 From the point of view of greater fairness, the five-constituency model has much to recommend it, but in terms of simplicity I am inclined to suggest that perhaps a three-constituency model would be best: North, West and Central-Eastern, with the Eastern division being simply included with the Central, as the name of the constituency would indicate. With a 54-member Parliament, the respective number of seats per constituency would be something like: Northern 9, Western 21, Central-Eastern 24. The Eastern Division might at first feel a bit aggrieved at being put in with the Central Division, but voters actually stand to gain from it. As a constituency by itself, the Eastern Division would only be entitled to 2 seats. Voters would have very little choice, and would certainly ‘waste’ their votes if they voted for a minor party, which clearly would have no chance of having a candidate elected. Many people originally from the Eastern Division have indeed moved into the Central Division. It seems highly likely they could have a representative elected there from out of the 24 candidates to be elected. For instance, although voting along strictly ethnic lines is not being encouraged, Rotumans would be quite likely to get a Rotuman candidate elected in the Central-Eastern constituency if Rotuma was a part of it, - (to have somebody also elected in the Western, or even Northern, Division would not be out of the question).
3.10 The question is: are constituencies of 24 and 21 candidates too big? My answer is: No! We are unused to multi-member constituencies in Fiji, but they operate in many countries of the world quite satisfactorily – even when each candidate is representing far higher numbers of voters than would be the case in Fiji. This sort of constituency is indeed new for Fiji, but if we are to improve our voting system, there is no point in shying away from anything that is new. To do that rigorously would, of course, be to stay with the old system! We need to have the courage to strike out into new areas, but not without foreseeing problems and taking steps to counteract them.
3.11 To go with the whole country as just one constituency would, I think, be too much again, as indicated earlier. The number of candidates for each voter to choose from would be very large, and the voter-candidate relationship, through a constituency, would be largely lost. To go with smaller constituencies, say as low as 5 candidates per constituency, would have serious consequences for proportionality. To compensate for this, it would be necessary to require that not all 54 seats could be voted for directly, but only, say, 46. The remaining 8 would need to be held aside to make up for serious deficiencies in proportional representation. This process itself would be complicated, requiring the totalling of party votes from all the constituencies to determine how the 8 seats would be assigned. Such extra complication would not recommend itself to the public.
3.12 Having such smaller constituencies would allow some of them at least to be coterminous with provinces (plus or minus certain districts). But that is probably not a good thing in Fiji’s current circumstances. These provinces or districts are often closely tied up with certain ‘vanua’ positions. Even if somebody in such a position does not himself run, indication of support for this or that candidate by such a person can make it difficult for his subjects to vote for someone else. There can thus be an unhealthy mixture of ‘vakavanua’ and political interests, whereas it would seem preferable to keep the two types of leadership distinct – both for the ‘itaukei’, who are directly affected, and for others too. Though there could still be some ‘vanua’ influence on the larger division-based constituencies proposed here, it would be much diluted by the number of provinces, with their different loyalties, that are voting together within the Division.
3.13 The proposed constituencies may seem large in area, but this has not been a problem in the past. The constituencies for General Voters (“Others”) over the years have been exactly of this size (and indeed the constituency for Rotumans extended over the whole country!). Campaigning, and then caring for their constituents, did not seem a problem for the small parties concerned. If it might seem difficult for a large party having numerous supporters, the party would have enough candidates in its ranks that it could divide up among them the task of campaigning in and caring for the constituency. A much more direct voter-candidate relationship could thus be maintained after all. The proposal of such large constituencies certainly seems worthy of adoption, especially since, as already mentioned, they provide good accountability to the voters.
4.0 Additional Observations
4.1 In the foregoing sections, the ideals of proportionality and fairness seem well taken care of. But what would be the situation regarding the other big issue desired by the Charter – the elimination of communal seats? While the ideal remains that people would vote for suitable citizens, regardless of ethnicity, it will clearly take a long time for this to be the practice throughout the country. Some areas of the country, e.g. Kadavu and Lau, have a very high percentage of ‘itaukei’ residents, while others, especially in the sugar belts, have a high percentage of Indo-Fijian residents. It is unrealistic to expect many of these people to vote for a person of another ethnicity when they rarely have occasion to mix. There is also the fact that racism cannot be banished by decree. Better mutual regard by Fiji’s various ethnic groups will take time to mature.
4.2 The elimination of communal seats, then, will not eliminate ethnically-based voting, nor will the Open List system force the issue. People will be free to vote for whom they please, and the parties to run whom they please. By using the Closed List system, some pressure could be prescribed in this regard. For instance, if parties were required to have a certain percentage of each ethnicity spread throughout their list (as distinct from having candidates of one ethnicity all placed at the beginning or end of the list), a multi-ethnic composition of Parliament, and indeed of all the major parties in it, would be assured. But the price to be paid for this is far too high. Firstly, the ethnicity of candidates would by definition become a major factor in compiling party lists. This would entirely sabotage the desire to see all people of the country regarded primarily as citizens rather than as of this or that ethnicity. Secondly, it is easy to envisage the situation where some candidates would not really be the people’s choice at all, and could hardly be said to represent them, because those candidates would have got elected only on the basis of having to be placed, by law, high on the party’s list. Thirdly, such manipulation of the party lists not only takes away the party’s freedom to put forward the most attractive list it can, but also seriously curtails voters’ freedom to have their votes count toward whom they really want to be elected.
4.3 The Open List system remains by far the most democratic and treats everybody equally as citizens. It makes no reference to ethnicity (or, for that matter, any other category). All is left up to the voters to do what they want with their votes. A party might seek to be multi-ethnic in nature and present an ethnically diverse list. But if the party is attractive mainly to people of one ethnicity, it could happen that the elected candidates are mainly of that ethnicity. This, I believe, is simply a fact of life that we must accept. There’s no point in putting forward people’s democratic right to vote for whom they please (one person, one vote, one value) and then get upset or even angry because they vote in a way we prefer they hadn’t. Democracy will not in all instances provide the very best answer. All we can do is set the parameters correctly, which the Open List does. After that, we can only hope that the good judgment of the people will provide a reasonable result. As stated above, a reduction of ethnic considerations in voting will take time to come about. Responsible campaigning by parties, however, can certainly contribute a lot to facilitating this move. We should not force the voters this way or that. It is time to treat them as adults, but try to draw from them genuinely well thought-out votes.
4.4 The point being made here is valid for more than just the ethnic question. Let’s take gender. Just as for ethnicity above, a party may decide to go for gender parity and run a list containing equal numbers of men and women. But if the voters are not on-side with this and if many of the women candidates, say, are not well-known, the nett effect of running so many women could be that the vote for women is divided among too many of them, such that fewer of them get elected than if a smaller number had been run in the first place.
4.5. It has become fashionable to speak of requiring a certain quota for women in Parliament – to require by law that a certain number of women be elected, or at least be on the party lists. Quotas should definitely not be employed. The Open List system provides people with plenty of freedom to decide whom they want to be elected. Parties will undoubtedly put women on their lists. Not to do so would be suicidal. Half the population are women after all, (although the last thing we want is for men to vote for men and women for women!). A party needs to consider carefully, though, just how many of any category of people it puts on its list. The great thing about the Open List system is that many categories of people can be put on the lists. While many people will consider the gender issue as important, there are other factors that come into play. Each voter has only one vote, so has to choose among a lot of competing values.
4.6 A compelling reason not to have quotas under Fiji’s circumstances comes from this very fact of competing values. If it’s okay to have quotas for women, then why not for ethnic groups? – and we end up going back to square one: communal seats! Then it will be: why not quotas for religious groups, such as Muslims, Catholics, etc. And why not for the handicapped, the elderly, youth, and so on? The great advantage of the Open List system is that it gives voters good leeway to choose among the various competing values. The freedom of the system needs to be exploited, not circumvented or prejudged. If there is some reasonably large group which seems to be regularly excluded, this would need to be investigated. A place for such a group might be considered for the Senate, (if such is retained in the new political arrangements).
4.7 These last paragraphs show that there is a mutual accountability taking place between parties and voters in the Open List system. Formerly, with single-member constituencies, parties would pick a candidate who they thought would have wide appeal (often a middle-aged male professional). With multi-member constituencies, parties are freer. Instead of choosing their candidates so that they are all of wide appeal, they are likely to do better if they choose several candidates who are of special appeal to certain sectors of the community (female, blue-collar, young, etc.). The parties put forward their candidates and platforms and challenge voters to vote for them. But to succeed, they cannot overplay their hand. While they may have certain goals deriving from their underlying ideology, they cannot afford to push too hard. Those parties will be most successful that gauge the mood of the people and, while challenging them perhaps in certain areas, basically respond to the direction the people want to go in. What’s more, the Open List provides better avenues for voters to keep parties to their promises. These things are what accountability is about, surely?
4.8 Because the accountability of the electoral system is basically to the people (see 1.1), there is good reason to have the primary features of the electoral system enshrined in the Constitution, rather than have too much important material as Electoral Law, which could be chopped and changed too readily by political parties for narrow party interests. This point has already been raised in 2.12. However, the 1997 Constitution provides an example of a situation where too much was tied down by the Constitution, which made it very difficult – in some cases almost impossible - to bring about desirable modifications. The two main features identified in 3.1 (having all seats open and adopting proportional representation) would best be included in the Constitution, and perhaps some other items too. But one doesn’t want vast tracts of the electoral process requiring constitutional change to have them amended. A robust and neutral process such as suggested in 2.12 would seem the best means to keep electoral processes up-to-date while making sure they are truly accountable to the people.