It is possible to map out a minimum platform for advancing the process of electoral reform. However reform cannot take place in a political vacuum.The context and the circumstances of electoral reform are extremely important. This is probably where sadly Govt may have been unwise both in its proposals and the timing of its introduction.Those opposed to Govt could marshall four very robust reasons to reject the Bill. Given the assurances by the Chairman of the Sachs Commission who is a very distinguished Constitutional Judge from South Africa, the Opposition not only submitted a very comprehensive paper and deponed before the Commission but also accepted the conclusions and recommendations of the electoral experts. Further, Model C of Sachs has been endorsed by the Collendavelloo Select Committee. But Govt has decided to reject the best formula of the Sachs Commission. Second, Govt is endorsing what has been comprehensively and categorically rejected by both the Sachs experts and by Collendavelloo. The Opposition could argue as to why the country should accept what is plainly a partisan solution spurned by experts. The attitude would have been different if Govt had proposed a modified version of the best formula of the experts (say Model C of Sachs but with fewer than 30 PR MPs). The Bill has been submitted by Govt to a world known electoral expert for his views. He has totally rejected it. Third, many would contend that it is not proper to introduce voting reform during an electoral campaign. Why wait one month before the dissolution of Parliament to introduce such important reform? The atmosphere is certainly not conducive to cross party discussion and for reaching consensus. Yet the Sachs report was submitted more than two years ago. And finally many could argue that Govt is concocting a scheme to obtain some seven seats in the next Parliament. As the Opposition is confident of a huge victory at the next elections,it could see the Bill as a ploy by Govt to secure a minimum presence in the House.

Notwithstanding the above caveats, I submit,hereunder, what I consider could be the main features of the reform in the light of what has been suggested by Government. It is certainly not fair to accept the constraints imposed by Government as electoral experts have proposed significantly better solutions. However if we remove some awful provisions from the Bill and include some best international practices in matters of electoral reform, I consider that it is possible to get the ball rolling. While it is certainly not the ideal solution, it will constitute a marked improvement on the current formula.

The two main constraints imposed by Government are an increase in the number of MPs by 10 and the maintenance of the 62 Find Past the Post (FPTP) seats. As argued by Sachs, they simply make the share of Proportional Representation (PR) seats in the total too low to ensure a good degree of fairness. But it can be intelligently managed if some changes are accepted. Government has also combined the element of proportionality with the objective of improved female representation.This poses an additional hurdle as they are usually handled differently. As the Prime Minister has stated, it would be only the beginning of the reform process.Changes will have to be brought later on in the light of the consequences and experience of the initial change.

Based on very extensive research on electoral systems in plural societies, consultations with international electoral experts and what obtains in other countries that have faced similar problems, my proposals are as follows :

i) Keep the 20 three member constituencies in the Island of Mauritius and the one two member riding in Rodrigues;

Then we choose between either of these two ways of introducing an element of proportionality and of providing for gender fairness.

ii) either subsume the 4 Best Loser seats into 18 PR seats. Basically we combine the 14 PR with the 4 Best Loser seats as proposed by Govt to give 62 FPTP and 18 PR seats in a Parliament of 80 MPs;

iii) or introduce 14 PR seats and retain the 4 Best Loser seats as suggested in the Bill,thus giving 62 FPTP, 14 PR and 4 Best Loser seats in a Parliament of 80 MPs.

If the major political parties present,as expected and as will be the case, a balanced and a diverse slate of constituency and PR candidates,the difference between (ii) and (iii) above would be marginal.

<B>A. Proposals in case of 62 FPTP and 18 PR seats</B>

i) the split between FPTP and PR seats would be 77.5 % FPTP and 22.5 % PR. The share of PR seats is low;

ii) to offset for this low number of PR MPs, the 18 PR seats should be alloted on a compensatory basis as recommended by Sachs,by Collendavelloo and by the international expert consulted by Govt. This would reduce the disproportionality between vote polled and seats obtained. For instance the MSM- RMM in 1995 would have obtained 18 seats for 20 % of votes while the LP-MMM would still have a very huge majority of 40 seats (60 minus 20). In 2000, the LP- PMXD would have secured, in total, 24 seats ( or 30% of seats) for 36.7% of votes while the MSM/ MMM would still command a large majority of 28 seats (54 minus 26).It will also avoid the 60-0 while being fair to the unsuccessful party;

iii) the 18 PR MPs should be returned from a closed and rank ordered Party list that must be published before elections as is the case in all countries and as recommended by Sachs, Collendavelloo and the international expert. Unlike the Best Loser system of returning PR MPs, Party List will not destabilise national unity and will not encourage racist, communal,and casteist electioneering. On the contrary the composition and rank of the party list will reflect the diversity and rainbowness of our nation;

iv) there should be two sets of votes as is the case in most countries (Germany, New Zealand, Wales,Scotland, Lesotho,etc) with mixed system. A first set of three votes in each constituency (except for 2 in Rodrigues) and a second vote for electors to choose their preferred party. Voter choice is one of the most important attributes of an electoral system. Two votes are essential to give choice to voters;

v) the allocation of PR seats should be based on Party vote.This is best international practice;

vi) we could use the D?Hondt formula to attribute seats. However the share of seats should be computed on the basis of eligible votes by deducting the votes of parties which have not reached the prescribed limit of 10%;

vii) there should be provision for double candidacies as is the case in practically all countries with a mixed system (Lesotho, Germany, New Zealand, etc). Elections will lose their panache if the top guns of political parties are not fielded in constituency contests. However they will be afforded a chance on the Party list also as they represent the ?heart and soul? of the party.It would simply be ridiculous to have the MMM, the Labour Party and the MSM in Parliament without their key players;

viii) My simulations show that a 5% threshold is adequate to keep communal parties out. However because of the lack of expertise and experience in Mauritius to assess the impact of threshold on party fragmentation, I realise the difficulty to convince even well intentioned people of the merits of this submission.I therefore leave it to the political stakeholders to sort it out, even if a compromise could be 7.5%;

ix) We should adopt best practice when a vacancy arises. Bye elections will continue to take place for constituency MPs.With respect to a vacancy for a PR parliamentarian,the next person on the party list should be appointed, provided that person is still a member of the eligible party;

x) There should be no crossing of the floor for PR MPs. Lesotho, the last country to have introduced electoral reform and which has extensively consulted international experts, does not allow PR MPs to cross the floor. The PR MP should resign in case he/she changes party loyalty. The argument is a very simple one. As the second vote is a Party one, the seat should remain with the eligible party. The problem is trickier for constituency MPs as there is also an element of personal vote. This explains why many countries have difficulties to handle defection by constituency MPs inpite of the vehement criticism against such practice.

<B>B. Way forward in case of 62 FPTP, 14 PR and 4 BL seats </B>

a) as above except that there would be only 14 PR seats. This would lower the share of PR seats to 17.5 % instead of 22.5 % ;

b) the same eligibility of 10 % (or 7.5 %) of national vote should apply for a party to be eligible for a Best Loser seat. This is essential for consistency purpose and to avoid the glaring anomaly of a national party with 9.9% not obtaining one PR seat while a single issue party with less then 1% of vote gets one Best Loser seat. The community of the candidate will be the same, however the party may be different;

c) there should be two sets of 2 Best loser seats. The first set of two will be allocated to the most successful unreturned candidates from the underrepresented community, irrespective of parties while the second set of two will be attributed so as to ensure that the will of the electorate is not altered.Similar to what exists today. This is important to avoid transforming winner into loser;

d) However there is a better alternative to (c) above. All four seats could be apportioned on the basis of both the underrepresented party and the underrepresented community. This has the double advantage of mitigating the underrepresentation of some communities while lowering the level of unfairness between parties. Obviously it will not change the majority. First we look at the underrepresented community, then we consider the underrepresented party. The ?right? person from the ?right? party gets the first seat. And likewise for the next three seats.

<B>C. Curing the democratic deficit of women underrepresentation</B>

Ideally we should reach the 30% female representation in 2005.The best solution would be as follows:

i) have at least one woman candidate in each of the 20 constituencies;

ii) provide for closed Party Lists to contain at least one third of female candidates;

iii) have a narrow ?zipper? so that candidates on Party lists are rank ordered in a sequence so that at least every third candidate is a woman;

However, I have to be both honest and practical. Based on current trends and the attitude of some stakeholders,the odds of Mauritius having a Parliament with 30 % female MPs in 2005 (a SADC commitment) is extremely long.To be fair,this may be attributable to seats being at a very high premium,thus leading to stiff competition for selection and to the difficulties of displacing incumbents.

One practical solution would be for political parties to agree on the target of 30% but to stagger its implementation over a period of ten years,spanning three elections,starting with 20% in 2005, rising to 25% in 2010 and reaching 30% in 2015.The 20% and 25% would apply to both the FPTP and the Party list.

<B>Concluding remark</B>

There are some timeless and universal principles that characterise electoral systems. However often history, context, political environment and the specific circumstances of the country will have a significant bearing on the choice of voting formula. Ideally Mauritius needs an electoral system with the following six main attributes. Stability, fairness, inclusion, voter choice, constituency links and avoidance of extremist parties. Only a mixed system as proposed above can achieve the best balance among these attributes.

The suggested pathway, even limited because of the constraints imposed by Govt, satisfies most, if not all, the characteristics of a good electoral system and is consistent with best international practice. The model prioritises stability with 77.5% or 82.5 % of seats returned by a FPTP mode. Fairness, though not total, is provided for by the compensatory formula to allot the 14 or 18 PR seats. Inclusion is guaranteed by enhanced female representation through the main door. Electors have greater choice with two sets of votes compared to only one set in the current system.The vital links between the electorate and their MPs are maintained with an overwhelming number of constituency MPs. In addition the smallness of the country will also ensure close contact between the PR MPπs and the voters. The very high threshold of 10 % of votes to be eligible for PR seats will certainly deny single issue parties access to Parliament.