Insights, perspectives and recommendations on balancing fairness, inclusion and representation with stability and effectiveness in the voting formula

Table of contents

Chapter 1 : An evaluation of the current system and of the proposed amendments

Chapter 2 : The best international practice of seat allocation in a mixed FPTP/PR system

Chapter 3 : A pathway of proposals to enhance stability in the voting formula of Rodrigues

Chapter 4 : Concluding remarks and the way forward

CHAPTER 1 : An evaluation of the current system and of the proposed amendments

1.1 The current electoral system

The current voting formula to elect members of the Rodrigues Regional Assembly is a version of a mixed FPTP/PR system. Out of a total of 18 seats, 12 are returned through a First Past The Post (FPTP) mode in 6 constituencies of 2 elected representatives each while the remaining 6 members are chosen from a pre-established Island wide party list using a compensatory PR algorithm. The ratio of the mixed system is 67 per cent FPTP and 33 per cent PR which is reasonable for devolved Assemblies that purport to balance the two key features of a good electoral system which are stability and governability on the one hand and fairness, inclusion and representation on the other. In a nutshell, the allotment of PR seats takes into account the disproportion between votes polled and seats obtained in the FPTP so that there is a better alignment between overall votes and seats in the Island to reflect the realities of party support. It ensures fair, inclusive and diverse political representation as a party with 45 per cent of the votes without any FPTP seat could obtain around 33 per cent of the total seats instead of nothing as could be the case in a pure FPTP system.

There are two sets of votes per elector with a threshold of 10 per cent to be eligible for the PR seats. Subsequently, a proviso was added to ensure that a party which wins at least seven FPTP seats (an absolute majority of the 12 FPTP seats) will, if required, be given additional PR seats to have an overall majority of 50 per cent of seats plus one. This is notwithstanding its share of the national vote. That clause was applied in 2012 to give three additional seats to OPR as it could not retain its majority after the apportionment of 4 and 2 PR seats to MR and FPR respectively. The final outcome was 11 seats for OPR, 8 for MR and 2 for FPR with an increase in the size of the Assembly from 18 to 21 members. It must be pointed out that it happened because the party with a majority of FPTP seats failed to capture a majority of the votes at 46.8 per cent in a three horse contest with all of them eligible for the PR seats.

1.2 The formula delivered as per expectation in Rodrigues in 2002

The 2002 results of the Rodrigues Regional Assembly elections were as anticipated by such a mixed FPTP/PR system. They are as follows

As OPR polled a majority of the votes with 54.97 per cent, it could maintain its overall majority after the allotment of the 6 PR seats. There is nothing puzzling that the absolute difference between the two parties shrinks after the allocation of the 6 PR seats from 4 to 2. This is indeed the very essence of the mixed proportional formula that compensates parties that have been disadvantaged in the FPTP race.

If the final margin were to stay the same as the FPTP difference in seats, there would have been no need for a different voting formula.

OPR took 67 per cent of the FPTP seats with 55 per cent of the votes (a positive degree of deviation of 12 percentage points) while MR captured 33 per cent of these seats with 44 per cent of the votes (a negative degree of deviation of 11 percentage points). OPR had thus a higher share of seats compared to its vote share and the reverse holds true for MR. With 54.97 per cent of the votes, the winner of the FPTP contest remains so after the allotment of the 6 PR seats, but the wedge between the two parties is narrowed from 4 to 2 seats. This is due to the relatively higher compensation received by the second party in the PR mode as it was somewhat “disadvantaged” in the FPTP mode. The degree of compensation depends on the percentage of votes polled by the eligible parties. The award of 4 PR seats to MR and 2 to OPR in 2002 reflects the twin facts that OPR won 8 of the 12 constituency seats with 55 per cent of the votes while MR captured 4 of these seats with 44 per cent. The share of the votes is key for the final allocation.

Had OPR polled 58 per cent of the votes and MR 41 per cent, the apportionment of PR seats would have been different with three to each party. Overall OPR would have obtained 11 seats and MR 7. Then the gap of 4 seats after the FPTP election would have remained unchanged after the application of the PR formula. On the other hand, had OPR taken 10 constituency seats with 56 per cent of the votes and MR only 2 such seats with 43 per cent of the votes, MR would have been entitled to all six PR seats. The overall total would have been the same as the actual results (10 OPR and 8 MR); only the split between the two sets of seats would be different. Under that scenario, the difference of 8 seats (10 OPR and 2 MR) after the FPTP mode would have come down to only 2 with the PR allocation (10 OPR and 8 MR). Nothing mysterious as it reflects the relative vote share of the two parties on the Island.

1.3 The system worked extremely well and produced a majority in 2006

The 2006 results as shown below is a strong rebuttal of the “majority turned minority” figleaf.

In this case, the PR formula helped to produce a majority government. And rightly so. Because of the geographical concentration of their support, both MR and OPR won the same number of FPTP seats with 6 elected representatives each even if MR polled 55.8 per cent of the votes while OPR took only 43.1 per cent of the votes, a significant difference of 12.7 percentage points. An unfair deadlocked FPTP outcome was resolved through the PR mode; as a result of its higher vote share, MR was allotted 4 PR seats and OPR 2 PR seats. Again MR was able to have a majority of seats because of its majority of votes at 55.8 per cent. Nothing untoward as the outcome is consistent with the expectation of the mixed formula.

1.4 The case of 2012 was unique with three parties eligible for PR seats and no party with a majority of vote

The results of the 2012 elections are shown below

The outcome was more complicated for OPR in 2012 than in 2002. It captured 67 per cent of FPTP seats with only 46.78 per cent of the votes while in 2002 it won the same number of FPTP seats on 54.97 per cent of the votes. Second it was a three cornered fight with all three eligible for the PR seats. The 6 PR seats were apportioned to MR and FPR in the proportion of 4 to 2 to reflect the reality of the votes. And after the allocation of the 6 PR seats, OPR lost its FPTP majority of 4. It became a hung Parliament with OPR and MR on 8 seats each and FPR on 2 seats. Using the special provision of guaranteeing a majority to a party that wins at least 7 FPTP seats, OPR was awarded an extra 3 seats to reach the threshold of 50 per cent plus 1 seat. OPR ended with 11 members out of an enhanced Assembly of 21 members instead of 18.

In 2002 and 2006, the system delivered a majority Government while being fair, inclusive and representative. One must understand that a mixed system will most of the time produce a more equitable result than a FPTP voting formula. The very purpose of a dose of PR is to correct the huge disproportion between the seat and vote shares of the FPTP system. The 2012 election was thrown off balance with three parties contesting the seats. Had there been only two parties eligible for the PR seats, OPR would have won 2 such seats and the MR would have kept its 4 PR seats. The overall result would have been 10 seats for OPR and 8 seats for MR, similar to what happened in 2002. It would have given a majority to OPR without using the special provision. Alternatively had OPR polled the same share of vote as in 2002, it would have gained a majority. With only 46.78 per cent of the votes, it did not win any PR seat to maintain its lead.

The KEY IS THE SHARE OF VOTES.

1.5 The unfairness of the most important criticism of the FPTP/PR formula

In a country used to only the FPTP voting formula, the single most important criticism of the mixed FPTP/PR electoral system of Rodrigues is that it alters the balance of power as it stands after the FPTP election. Little do some people realise that the very purpose of introducing a dose of PR is to precisely correct the anomalies of the FPTP which produces a situation where “winner takes all” or there is huge distortion between vote share and seat share as is very often the case in Mauritius. Many people contest the fact that in 2002, OPR took 8 of the 12 FPTP seats giving it a margin of 4 members but with the allocation of the 6 PR seats and taking into account the vote share of MR, OPR ended up with 10 seats compared to 8 for the MR, thus yielding a reduced majority of 2 seats only.

The predicament was presumably worse in 2012 when OPR won 8 of the 12 FPTP seats and did not secure any of the six PR seats as it took only 46.7 per cent of the votes and had already captured 67 per cent of the seats. On paper, an alleged victory was converted into a hung Assembly. Worse the combined forces of MR and FPR could have formed a Government at the expense of OPR. The special provision was applied and three additional seats allotted to OPR. The probability of this reversal happening is quite high when the party that wins an absolute majority of FPTP seats polls less that 50 per cent of the Island vote.

Another criticism of the voting formula is that the very thin majority leaves the system open to the unacceptable practices of crossing of the floor that could threaten the stability of government. However one should acknowledge that even in a pure FPTP formula, the majority could be so fragile as to lead to floor crossing. We have seen it in Mauritius with very low majority in 1976-82 Parliament. There are solutions to address this problem.

The polarisation of local politics in Rodrigues is such that elections are closely fought contests. This has been the case in all three elections held since 2002. It translates into a very small seat advantage in an Assembly of 18 members once PR seats are apportioned. The difference in vote between OPR and MR, either at the constituency or the Island level, was closer in the 2012 election. With a small change in voting pattern, the FPTP result could have been different from the 8-4 in favour of OPR. Some may argue that this is a case for all parties to work together in the best interests of Rodrigues.

1.6 Expressing surprise when the results are clearly what is intended

There is one simple truism which unfortunately many well intentioned people seem to forget. Results under a mixed FPTP/PR formula are less unfair than under FPTP system. This is the very purpose of introducing a dose of PR in the FPTP system. There is much better correlation between votes and seats under a mixed system than under a FPTP formula. Yet, at times, the very people who advocate reform to cure the unfairness and injustices of the FPTP mode seem to show concern when the allocation of the PR seats in a mixed system does actually lower the margin of victory of the party that emerges first in the FPTP election. Why on earth should we introduce a dose of PR if the outcome of the mixed FPTP/PR system were to be the same as that of a pure FPTP as argued by some?

We must be internally consistent. We cannot change a FPTP system that is unfair in order to make it more representative and then be surprised when the final outcome is a fairer one in terms of the alignment between votes and seats!

This very simple point was extremely well articulated by the Select Committee of 2002 chaired by Collendavelloo:

“At the end of the day, we need to know what we want. It does appear incongruous to clamour against the unfairness of the present system and we agree that we require a dose of PR to mitigate the rigours of the system. We ought not therefore to be surprised when a dose of PR precisely achieves this result.”

1.7 Minority becomes majority upended

Some political actors seem to have a strange definition of who should be the winner in a mixed FPTP/PR system. It is patently incorrect to argue that the party that takes a majority of FPTP seats SHOULD ALWAYS be the overall winner and that any change in the outcome after the allocation of the PR seats is tantamount to distorting the result and to “converting a majority into a minority”. It is the very opposite that is true. The PR mode precisely aims at preventing A MINORITY FROM BECOMING A MAJORITY by the artificial creation of a manufactured majority unrelated to the share of vote polled by parties.

Again this was very well canvassed by Collendavelloo in the 2002 Select Committee report:

“There is no conversion of a “majority” into a minority since the final results only serve to proclaim the REALITY OF THE VOTE and not the contingency of an unfair system where the minority becomes the majority”.

This was spot on. What happens should depend on the share of votes polled by parties.

The result of the election of 2015 in the UK constituency of South Belfast clearly shows the extreme perversion of a formula that guarantees the FPTP winner to be the overall winner in a mixed FPTP/PR system.

As it is a FPTP system, the winner is Mc Donnell of SDLP party who polled only 24.5 per cent of the votes in a multi-party contest, even if 75.5 per cent of the electorate voted against him. Assume this outcome is replicated in a few constituencies. The question is whether SDLP should be guaranteed to remain the overall winner if there were PR seats in the system. If the SDLP with only 24.5 per cent of votes were given additional PR seats to retain a majority in a mixed system, we would have converted a minority into a majority. This would be an awful system. However should the party take around 50 per cent of the votes, there would be no problem to have a majority.

1.8 The proposed amendments to eliminate the PR component of the mixed system

While there is no perfect voting formula, it is clear that the amendments being crafted are deeply flawed, extremely unfair and will distort political inclusion and representativeness in the Island. They are utterly regressive as they simply eliminate the PR mode of the mixed formula and replace it with a form of Best Loser. While an element of specific context is important in the design of an electoral system, it must be noted that what is being envisaged does not exist anywhere, the more so in any devolved Parliament that uses a mixed member system to elect representatives.

It is proposed to change the current formula by doing away with the PR mode and replacing it with a system that is akin to a Best Loser where both successful and unsuccessful parties share the additional seats based on the number of FPTP members returned. There will hardly be any relationship between the share of votes and the share of seats of parties as is presently the case. 18 members would be returned in 6 three member ridings and 4 additional seats would be distributed to parties that win FPTP seats. The allotment of the 4 additional seats is based NOT ON THE SHARE OF VOTES but on the number of seats won in the FPTP mode.

There is provision for different outcomes in the allocation of the 4 Best Loser seats contingent on the result of the FPTP election. I shall give four illustrations to explain the working of the proposed system:

1. if a winning party takes more than 15 of the 18 FPTP seats, it shall obtain 1 additional seat while the three remaining seats shall go to the losing party with the highest number of FPTP seats;
    
2. if both parties capture 9 seats each, the party with the higher share of votes (even if it is by 1 vote) will be allocated all of the four additional seats;
    
3. in other cases – between 10 seats for one party and 8 seats for the other and between 14 seats for one party and 4 seats for the other, both will obtain 2 additional seats each;
    
4. as a losing party is rewarded only if it wins at least a FPTP seat, there shall be no allotment of seat to the losing party in case of a 18-0. Unless the clause is amended to provide for this special case and to award some of the additional seats to the losing party.

There is provision to deal with cases where there are more than two unsuccessful parties.

In a three cornered contest, if one party takes all the 18 seats on only 45 per cent of the vote, it will be allocated 1 out of the 4 additional seats and it is not clear (at least in the first version of the draft where a losing party is defined as a party other than the winning party which has won the largest number of seats at the FPTP election) whether any seats will be allotted to parties that have polled even 40 per cent of the votes without returning a single FPTP member.

We are thus in a situation of 19-0.

I understand a change may be made to distribute the 3 remaining seats to unsuccessful parties. Even this would leave the proposed amendments as highly regressive. Essentially it will make the electoral system of Rodrigues far less representative of the wish of the people and more unfair in terms of the vote-seat ratio. It will lead to huge distortion as it is a mirror image what we currently have in Mauritius. It is indeed very ironical. While we want to change the voting formula in Mauritius because of its unfairness, we are doing exactly the reverse in Rodrigues by introducing a very inequitable electoral system.

 1.9 The highly flawed process to introduce reforms

Changing the rules of converting votes into seats is often seen as an abstruse political game theory, driven by the purely strategic interests of parties. The consideration of political advantage may be a key factor in modifying the electoral system. Often narrow partisan concerns take precedence over broader national interests. Parties often embrace strategies that would maximise their chances of converting votes into seats in Parliament. It is pure politics, sheer electoral strategy and plain tactical considerations. We have a duty and a responsibility to avoid such temptations; otherwise the voting formula will be subject to frequent changes depending on the interests of political stakeholders.

Electoral system should not be lightly altered to satisfy the short term whims and caprices of some political actors, nor should it be based only on current context and circumstances following the outcome of one election only. Rather it should fulfill the long term democratic aspirations of the people of Rodrigues in terms of fair, inclusive and equitable representation while providing for stability and governability.

The amendments seem to fail many tests of good electoral reforms even in terms of processes.

First there has been no broad consultations, no informed debate, no detailed discussions, no cross country comparisons and no impact analysis as was the case when Judge Ahnee recommended the current formula;

Second there is no expert report that critically analyses the current situation and makes a compelling case for reforming the system after analysing various options as it happened with the Sachs commission;

Third, it looks like a knee jerk over reaction to meet only short term objectives after the results of the 2012 elections without any regard for the long term sustainability of the system;

Fourth, it fails to acknowledge that the electoral system worked well in 2002 and produced a majority in 2006;

Fifth, the rules of the game are being changed on the eve of the election. This is hardly done in a robust democracy.

Sixth, it gives the clear impression that it is designed to promote partisan interests. It could very well lead to a situation where one party may take most, if not all, of the seats with only around 45 per cent of the Island vote in a robust three cornered fight.

1.10 A preliminary evaluation of the proposed changes

Only the main elements of the new formula are being distilled as the text has not been made public yet. It may still be work in progress as far as some details are concerned. This assessment is based on information available even if some initial recommendations may be subject to modifications. However the essence and the contours of the new electoral architecture are known:

1. the share of the FPTP seats is increasing considerably at the expense of the other mode. Currently there are 12 FPTP seats out of a total of 18 seats. That number will rise to 18 out of a total of 22 with each of the 6 region returning 3 members. The percentage of FPTP thus rises from 67 per cent to 82 per cent . The chances of having extremely unfair results will grow considerably as is the case in all FPTP voting systems ;
    
2. the PR component of the mixed system which is so important to ensure fairness, inclusion and representativeness is being axed. Henceforth there will be no compensation to parties that receive a sizeable share of island votes but fail to win FPTP seats. The balance between governability and fairness is being thrown overboard in favour of the former;
    
3. the four additional seats shall be allotted on the basis of seats won at the FPTP contest and not on the share of votes polled by each party. The PR mode is thus removed and replaced by a Best Loser variant;
    
4. we are opening the door to huge distortions between vote shares and seat shares. In the current system, a party that polls 45 per cent of votes is likely to obtain around 33 per cent of the seats if it does not capture any of the 12 FPTP seats while in the proposed formula, it may end up with nothing at all or at best only 13 per cent of seats (3 additional seats out of 22). This is most unfair ;
    
5. we are introducing the extreme perversions of FPTP in Rodrigues. In a three party contest (as is likely to be the case), the predicament can be dramatic in terms of electoral unfairness as a party can take all the 18 FPTP seats on only 40 per cent of the votes. To add insult to injury it will also secure at least one of the remaining 4 additional seats. The remaining 60 per cent of the electorate of Rodrigues will either not be represented at all (if seats won at the FPTP election is used to allot the four seats) or at best obtain 13 per cent of the 22 seats (only if 3 of the 4 additional seats are allocated to the unsuccessful parties);
    
6. we are giving up the prospect of a tie in seats being resolved by the allocation of PR seats on the basis of VOTE SHARE in a fair manner. In view of the concentration of support for the two main parties in Rodrigues, it is possible to have an equal number of FPTP seats with vastly different shares of vote. With the proposed formula, the party with the higher share of vote – even if it is one – is allotted all of the 4 additional seats. This is not fair as it should reflect the realities of their respective vote shares. In the current FPTP/PR formula, the party that has a higher percentage of votes is likely to be the overall winner without being allocated all additional seats. It in fact occurred in 2006 as shown below
   
    
   

   	<p>&nbsp;</p>
   
   	<p style="margin-left:14.2pt"><br />
   		MR and OPR took 6 seats each even if MR polled 55.8 per cent of the votes and OPR only 43.1 per cent. As the PR mode of the mixed system also takes into account the share of vote, MR emerged as the eventual overall winner with 4 PR seats against 2 for OPR. This in fact reflected the wish of the electorate of the Island. Had the proposed system been in place, MR would have obtained all 4 seats while it should be distributed on the basis of their respective vote shares. It is crucial for the additional seats to be distributed on the share of votes if it is to reflect the reality of the vote;</p>
   </li>
   <li>with the new formula, there is a real danger of a minority party winning the election at the expense of a party with a higher share of votes. This would be an absolute calamity for Rodrigues. In 2006, had 113 people cast their votes differently in the Port-Mathurin district, Francisco François of OPR would have won the second seat at the expense of Louis Perrine of MR. Overall the OPR would have taken 7 FPTP seats with 43.1 per cent of the vote and MR only 5 FPTP seats with 55.8 per cent. A party with a minority of vote would have been converted into one with a majority in seats because of the anomaly of an electoral system. While the current mixed FPTP/PR formula would cure this ailment and restore the primacy of the party with more votes, the proposed FPTP/Best Loser will amplify it. In the current system, the compensatory formula would remedy that anomaly with 5 PR seats to MR and one to OPR to reflect the higher vote share of MR. The final outcome would have been 10 seats for MR and 8 for OPR thus respecting the wish of the electorate. However, the proposed amendment would give two seats to each party without recognising the higher vote share of MR. We would thus end up with the minority party still in the lead by 2 seats after the allocation of the additional 4 seats. This is in fact &ldquo;minority turned majority by the perversion of a voting formula&rdquo;;<br />
   	&nbsp;</li>
   <li>The increase from 18 to 22 seats is significant for such a small electorate and an Assembly that meets so rarely. There is close to 24,000 voters in Rodrigues with an average number of electors per local region at around 4,000. The number of seats are simply too high for such a small electorate. The more so that the Assembly meets only once every three months. Scarce financial resources can be much better utilised for the welfare of our compatriots of Rodrigues than such a high rise in the number of Assembly members. Worse, it does not even serve the purpose of reflecting the reality of the vote on the Island.</li>
   

1.11 Beware of many possible perversions

The proposed amendments are likely to expose Rodrigues to electoral anomalies and perversions.

i) Case 1: a highly disproportionate result in a two party contest
 

Party A captures all 18 FPTP seats on 54 per cent of the vote while Party B has no seat on 46 per cent of the votes. In the proposed system, either the party with 46 per cent of votes has nothing (as it has not won any FPTP seat to be eligible for the 4 additional seats) or it obtains only 13.6 per cent of seats if there is an amendment to allocate the remaining 3 additional seats to the losing party. It is either a 19-0 or a very unfair outcome of 19-3 for the party with 46 per cent of the votes.

In the current formula with 12 FPTP and 6 PR seats, the party with 46 per cent of the votes would end up with 6 PR seats, representing 33 per cent of the Assembly seats. The final outcome would be as follows

Party A with 12 seats (67 per cent) on 54 per cent of the votes

Part B with 6 seats (33 per cent) on 46 per cent of the votes

That is a much fairer outcome.

ii) Case 2: an extremely flawed outcome in a three party contest

Here the situation is more dramatic

Party A takes all the FPTP seats on only 44 per cent of the votes while Party B and Party C have no seat with 40 per cent and 16 per cent of the votes respectively.

If the distribution of the 4 additional seats were contingent on the losing party taking at least one FPTP seat, it would be a calamity as neither Party B nor Party C with a combined vote of 56 per cent would qualify for even one Best Loser seat.

If there is a modification, the losing parties will collect a maximum of 3 seats only. They would have a miserly 13.6 per cent of seats with 56 per cent of the votes.

In the current system, the two losing parties would share all of the 6 PR seats.

The outcome is much fairer in the current formula in terms of the share of seats to the share of vote.

iii) Case 3 : the nightmare scenario with loser becoming winner and winner turned loser

It is a disaster of gigantic proportion as we are in the worst possible nightmare.

This could easily happen in Rodrigues because of the concentration of political support on a regional basis. It nearly happened in 2006.

Party B takes 10 of the FPTP seats on 43.9 per cent of the votes while Party A could secure only 8 FPTP seats on 56.1 per cent of the votes;

The proposed amendment would allocate 2 seats to both parties.

The final outcome would be as follows:

Party B would obtain a total of 12 seats on only 43.9 per cent of the votes

Party A would secure a total of 10 seats on 56.1 per cent of the votes

Party B, a significant loser in terms of vote would be the winner.

In the current system of 12 FPTP and 6 PR seats, this cannot happen.

As Party A has polled a significantly higher of vote at 56.1 per cent, it will be awarded all the PR seats

The final outcome would be such that Party A with 56.1 per cent of the vote would end up as the overall winner.

CHAPTER 2 : The best international practice of seat allocation in a mixed FPTP/PR system

2.1 The split of seats between FPTP and PR modes in mixed electoral systems

There are increasingly more countries embracing mixed FPTP/PR electoral systems so as to ensure a balance between stability and fairness. The rule is very simple to ascertain the weight of each of the two modes in the mixed formula. All things constant, the more FPTP seats a mixed system has, the more stable it tends to be, while the more PR seats it features, the fairer it is.

The distribution of seats between FPTP and PR in mixed electoral systems varies across countries. The table below gives the split between the FPTP and the PR components in 25 countries that have such mixed formulae.

The variation is significant with the Democratic Republic of Congo at 88 per cent and South Korea at 18 per cent of PR seats. Countries such as Germany, Lithuania and Ukraine divide it equally between the two modes. Japan has 60 per cent FPTP and 40 per cent PR while in New Zealand 58 per cent of MPs are returned through FPTP and 42 per cent along PR, similar to what obtains in Scotland, Bolivia and the London Assembly. Quite a few countries have one third of total seats chosen from the PR mode, similar to what obtains in Rodrigues currently.

 Split between FPTP and PR seats in mixed system

2.2 How is winner determined in mixed FPTP/PR systems ? : Some case studies across the world

Even if it should be quite straightforward, there is disconcerting confusion in the minds of many political actors on the basis of determining the final outcome in elections held under mixed FPTP/PR formulae. It is similar to a game of football. The winner is never the team that is ahead after the first half but the one that wins after the second half. It is best international practice in countries with a mixed FPTP/PR formula to declare the winner only after the allocation of BOTH FPTP and PR seats as it guarantees that MINORITY DOES NOT BECOME MAJORITY. This is the case from Germany to New Zealand, from Scotland to Wales, from Lesotho to the London Assembly. It is not even a subject of debate among political scientists, electoral practitioners or policy makers as it is plain that this is the way to carry out such distribution of seats.

IT IS BASED ON THE SHARE OF VOTES OF EACH PARTY

2.3 The German mixed FPTP/PR model

Germany has a mixed FPTP/PR system with an equal share of the two modes. However because the allocation is carried out on a regional basis, there are often some “overhang” seats which slightly change the ratio between FPTP and PR seats.

 The result of the last election of 2013 in Germany is shown below

The allocation of the PR seats is based on a formula which is quite similar to the algorithm used in Rodrigues. The CDU/CSU took 236 out of the 299 FPTP seats (79 per cent), a massive absolute majority of 173 seats in the first half of the game. However, as it polled far less than 50 per cent of the popular vote (only 41.5 per cent), it could not retain its majority after the allocation of the PR seats. It secured only 311 seats out of a total of 631. It had no choice than to enter into a coalition with the SDP to govern. Mrs Merkel neither protested about “alleged FPTP majority converted into overall minority” nor called for a change in the electoral system of Germany as the rules are well established. Had the CDU/CSU captured a higher share of the popular vote, it would have obtained an overall majority.

2.4 The mixed FPTP/PR formula of the Scottish Parliament

The Scottish Parliament is returned through a mixed FPTP/PR model with 73 FPTP and 56 PR seats. The ratio is 57 per cent for FPTP and 43 per cent for PR.

The results of the 2016 elections are shown in the table

The SNP took 59 of the 73 FPTP seats, yielding a thumping majority of 45 MPs. However, with only a minority of vote on 41.7 per cent, it could not maintain its overall lead. It ended short of a majority of 2 seats. The outcome simply mirrors the reality of the votes. Nicola Sturgeon has not asked for a change of the formula or cried foul about a “majority converted into a minority”.

It is simply not true!

2.5 The mixed FPTP/PR system of the Welsh Assembly

The Regional Assembly of Wales has 40 FPTP and 20 PR members. The share is similar to Rodrigues at 67 per cent for FPTP and 33 per cent for PR seats. The allocation of PR seats is close to what exists in Rodrigues with a compensatory formula.

The outcome of the last election held in 2016 are given below

Labour won 27 out of the 40 FPTP seats, a majority of 14 members. However as it polled only 31.5 per cent (again a minority of vote), it could not keep its majority after the allocation of the 20 PR seats. It ended up with an overall tally of 48 out of 100 members.

2.6 The mixed FPTP/PR of the London Assembly

The London Assembly is composed of 14 FPTP and 11 PR seats giving a ratio of 56 per cent of FPTP seats and 44 per cent of PR members. Again the PR seats are distributed in a similar method to that adopted in Rodrigues.

The results of the 2016 elections where Sadiq Khan was elected Mayor of London are shown below.

Labour captured 9 of the 14 FPTP seats on only 39.9 per cent of the votes. It had a majority of 4 FPTP seats. But as its share of votes was low, it could not hang on to its majority after the PR allotment. It finished with 12 out of 25 seats. Mayor Khan did not call for a change in the system and did not describe it as a “majority converted into a minority”.

2.7 The mixed FPTP/PR model in New Zealand

As a result of the unfairness of the FPTP system, New Zealand changed its voting formula to a mixed FPTP/PR model. It has 71 FPTP and 50 PR seats. It means that FPTP represents 59 percent and PR 41 per cent of the overall number of seats. It allots the PR seats in a way akin to what is done in Rodrigues so as to ensure fairness of representation. The results of the 2014 election give a clear indication of how the winner is determined in such an advanced democracy.

The National Party won 41 of the 71 FPTP seats giving it a majority of 11 Mps in the first round. However, as it is almost a full proportional system, the winner of the FPTP contest could not retain its majority as it polled less than 50 per cent of the national vote (it took 47.04 per cent). It was short of a majority by 1 seat. However it formed the government as it had the largest number of MPs.

I sincerely hope these live examples from mature democracies will put an end to this wrong debate about who should be the winner in a mixed electoral system.

IT IS THE SHARE OF VOTE THAT DETERMINES WHO IS THE ULTIMATE WINNER AND NOT THE PARTY WITH A MAJORITY OF FPTP SEATS.

Of course the two could be the same depending on the share of vote polled by the party with a majority of FPTP seats. The higher its share of vote, the more likely it will hold on to the majority.

The lower its vote share, the less likely that it will retain it.

It should be pointed out that in none of these countries is there a special provision to guarantee a party an overall majority simply because it has won a majority of FPTP seats.

However in all five cases, the party with the most FPTP seats still has a premium of seats. For instance in Germany the CDU/CSU took 49.3 per cent of seats on 41.5 per cent of votes, in Scotland the SNP captured 49 per cent of overall seats with 41.7 per cent of votes. In Wales, Labour won 48 per cent of seats on 31.8 per cent of votes while at the London Assembly, Labour gained 48 per cent of all seats on 39.9 per cent of votes. In New Zealand, the National Party obtained 49.6 per cent of seats on 47.04 per cent of votes.

However in all cases, the vote share was not high enough to secure an overall majority after the distribution of the PR seats.

2.8 The mixed FPTP/PR Lesotho formula

 Lesotho, a former British colony and a SADC country that had a FPTP formula, adopted a mixed FPTP/PR system in 2002 because of the unfairness of the FPTP. It has a combination of 80 FPTP and 40 PR seats, similar in percentage to what exists in Rodrigues (67 per cent for FPTP and 33 per cent for PR) .The PR seats are apportioned on the same principle as in Rodrigues. Therefore, the share of vote of the party winning a majority of FPTP seats is crucial to determine whether it retains its lead after the distribution of the PR seats.

The results of both the 2002 and 2012 elections in Lesotho below underscore very clearly how the share of votes matters a lot.

In 2002, the party that took almost all of the FPTP seats polled 54.6 per cent of the votes. It was thus able to maintain its lead after the allotment of the PR seats even if it did not receive any of them.

However in 2012, DC won a majority of FPTP seats on only 39.6 per cent of the votes (far less than 50 per cent). Under these circumstances, it could not retain its majority after the allotment of the PR seats. It ended up with 48 out of the 120 members as its share of vote was too low at 39.6 per cent.

2.9 The mixed system does not always require 50 % of vote to retain majority

While a share of 50 per cent will usually guarantee that the party with a majority of FPTP keeps its lead after the distribution of the PR seats, there are cases where close to 45 per cent of the popular vote may be enough to accomplish that task.

It occurred in the Scottish Assembly elections in 2011.

With 44.04 per cent (higher that its 41.7 per cent of 2016), the SNP captured 53 of the 73 FPTP, a sizeable majority of 33 members. It was able to secure an overall reduced majority of 9 after the allocation of the PR seats.

CHAPTER 3 :A pathway of proposals to enhance stability in the voting formula of Rodrigues

 3.1 Do not throw the baby with the bathwater

We need to draw the lessons of the last three elections in Rodrigues and make the necessary amendments to ensure the proper functioning of the Assembly. We should be very careful not to throw the baby with the bathwater. The dilemma will always be between a system that produces a clear majority even if it is very unfair as is the case with the proposed amendments and one that is fair and equitable to parties that poll a given percentage of votes. Representation of the different shades of opinion is crucial in a representative democracy. There is an absolute need to ensure diversity and inclusion of political parties, especially in a very small society as Rodrigues. The challenge is to find a compromise between stability and governability on the one hand and fairness, inclusion and representation on the other.

Even if it was dealing with the electoral system of Mauritius, the Select Committee of 2002 chaired by Collendavelloo aptly describes the challenges we face in Rodrigues when it stated

“there is unanimity that a measure of proportional representation is necessary so as to ensure that our electoral system becomes fairer, more democratic, more representative and gives an adequate opportunity for those who vote against Government to have their voices heard in Parliament. At the same time, we need to ensure that we retain the degree of stability which will ensure stable and efficient government as part of the democratic process”.

We need a balanced and a composed approach and not an abrupt shock that will severely disturb the balance between stability and fairness.

If there are too many PR seats, the stability of the system is likely to be challenged and it might be difficult either to have a clear winner or for the winner to govern for the duration of the mandate. However, if there are too few PR seats, it might be hard to lower the disproportion of the FPTP and there will be no reasonable correspondence between voter support for a party and the number of seats it actually wins. This has the potential of converting the system into an “elective dictatorship”.

With those twin objectives in mind, there are various options available to policy makers with respect to changes that could be brought to the voting formula of the Rodrigues Regional Assembly.

3.2 One preferred option and two second best options

Option 1: keep the mixed FPTP/PR formula but make some changes to enhance its stability features.

 I submit this is the preferred option as it will reflect the reality of the vote on the ground.

However, if this is unacceptable to some policy makers, in spite of its largely superior characteristics for democracy, we have a responsibility to offer some better alternatives to prevent the disastrous amendments being comtemplated. Essentially to choose between the lesser of the evils.

Option 2: adopt a new formula that guarantees a fair share of seats to unsuccessful parties while ensuring that the FPTP clear winner retains its overall majority subject to some conditions in terms of vote share;

Option 3: keep the mixed FPTP/PR but change the formula so that FPTP members are elected with at least 50 per cent of the vote.

3.3 Option 1 : keep the current FPTP/PR formula with some modifications

Recommendation 1

It is recommended that the special provision which “manufactures an artificial majority” be removed.

We simply cannot guarantee a cooked majority with less than a threshold of XX per cent (say, between 45 to 50 per cent) of votes. In effect, the current clause does not make the crucial distinction between a party that wins 7 seats with 55 per cent of votes in a two horse contest and one that captures 7 seats with only 30 per cent of the votes in a multi-party election. In one election in Papua New Guinea, a party took a majority of FPTP seats on less than 20 per cent of the votes as there were more than 6 parties contesting the election.

How on earth can we guarantee an absolute majority in the Assembly to such a party? This would be the perfect definition of a minority converted into a majority by the quirkiness of an electoral formula.

Unsurprisingly, this flawed proviso does not exist in any democratic country with mixed systems. One can only hope that those who conceived this “unheard of” amendment fully appreciate its political implications for Rodrigues. What would happen if a party with 52 per cent of the vote is deprived of a majority of seats through an electoral trickery? It is a recipe for political riots and civil unrests. I strongly believe that this highly reprehensible clause should be scrapped forthwith. As both votes are equally important, the process should go its full length in order to determine who wins more seats, as is the case in all countries with mixed systems. We need to understand that it is a game with two halves. Otherwise we are opening the gates to hell.

The above anomaly takes an even more alarming proportion in a multi party system. To win a FPTP constituency, one requires only a plurality and not a majority of votes. As we had practically only two parties contesting the Rodrigues polls in 2002 and 2006, all elected candidates garnered more than 50 per cent of vote. However, in a three or four cornered contest, only a plurality of votes is necessary to triumph as was the case in 2012.

 In the 2015 elections in Canada, a Bloc Quebecois candidate managed to win his riding with just 28.6 per cent of voters casting a ballot for him. Assume, for argument sake, that in a multi party contest, four of the six constituencies in Rodrigues produce such a result. It would give one party 8 FPTP seats with only 28.6 per cent of the votes. The lower the winning candidate’s share of the vote, the less the result can be said to reflect the will of the island electorate and the more the outcome is a lottery rather than a serious election. The application of the last minute proviso would allow that party to artificially obtain enough PR seats to command an overall majority of 50 per cent plus one seat. Nobody would accept such outrageous results. It would cause pandemonium, provoke riots, lead to a quick descent to hell. It would be impossible to govern.

Recommendation 2

Should a special provision be required to prevent a “majority from being converted into a minority”, it is recommended that it must be very different from the current one

A majority is a majority of votes in the population and not a majority of seats through the contingencies of a voting formula.

It is possible in a FPTP system for a party that polls 50 per cent or more of the popular vote to lose the election in terms of seats. It does happen sometime in countries with such voting formula.

It nearly occurred in Rodrigues in 2006. New Zealand changed its electoral system precisely for that very reason.

To prevent such a disastrous anomaly, there could be a provision where if a party with fifty per cent or more of the Island vote does not have a FPTP majority, additional seats will be given to that party to ensure that it has a majority of seats in the Assembly.

Assume in the system being proposed that Party A takes 10 FPTP seats on 45 per cent of the Island vote while Party B obtains only 8 FPTP seats on 55 per cent of vote.

This is a very flawed outcome.

The special provision would then grant enough seats to Party B so that it has an overall majority after the allocation of the additional seats.

In that scenario, all four seats will go to Party B which would end up with 12 seats to 10 seats for Party A.

Recommendation 3

It is recommended that we consider altering the split between the number of FPTP and PR seats from the current 12 FPTP and 6 PR seats to 12 FPTP and 5 PR seats.

While I am against an artificial provision that could give a 50 per cent plus 1 seat to a party that has polled far lower than 40 per cent of the votes, I am fully aware of the crucial importance of having a formula that helps to produce a stable and cohesive government. One possible solution would be to change the distribution of seats between the two modes. Currently 12 seats or 66.7 % of the total number are chosen through a FPTP mode and 6 seats or 33.3 % through a PR one. The higher the share of FPTP seats and the lower the percentage of PR seats, the more likely the system will produce a clear-cut result.

We could retain the six two member constituencies and lower the number of PR seats from 6 to 5.

This has two distinct advantages:

1. first, it increases the share of FPTP seats from 66.7 per cent to 70.6 per cent while it lowers the percentage of PR seats from 33.3 per cent to 29.4 per cent thus giving a fillip to stability;
    
2. second, the odd number of seats is more conducive to producing a working majority.

The suggestion to raise the number of districts from 6 to 7 will lead to protracted dispute of electoral boundaries especially in the light of the heavy concentration of geographical support for the two main parties. There will be acrimonious debate as is often the case in Mauritius.

Other combinations are possible to enhance the prospects of a stable government. However it has to be a balancing act as the more we move in the direction of FPTP seats, the higher will be the unfairness and the lack of inclusion of the voting formula as represented by the disproportion between the share of votes and the share of seats.

For instance, in addition to being excessive in the number of seats for a small Assembly that meets very rarely, the proposal to have 18 FPTP members in 6 districts of 3 members each and 6 Island Wide PR will tilt the balance too much against fairness, inclusion and representation. A party with 47 per cent of vote will end up with only 25 per cent of seats compared to the current 33 per cent. Also the risks of a repeat of the 2012 scenario does not go away in a three cornered contest.

Recommendation 4

It is recommended that we consider having only one set of vote and to use the same vote for allocating the PR seats.

This will enhance the stability feature of the system.

There are currently two sets of votes for the RRA. Each elector has two votes in each FPTP district to return 2 elected members. In addition the elector also casts a national vote for a party of his/her choice and this vote determines the allocation of the PR seats.

With 2 sets of votes – one at the constituency level and another Island-wide – the elector can split the vote by voting for one party at the constituency level and for another one at the Island level. This could affect the allotment of the PR seats. It is perfectly possible to use the votes cast for constituency candidates to choose the PR elected members. It can make a difference in favour of stability.

For instance in the 2012 elections, the split vote worked in favour of FPR. With 2 sets of votes, FPR tallied 10.7 per cent of the votes at the Island level and thus met the 10 per cent threshold to be eligible for PR seats. It obtained 2 PR seats as 10.7 per cent of 18 rounds up to 2 members. The compensatory PR exercise also allotted 4 PR seats to MR. The final outcome was 8 seats each for OPR and MR and 2 for FPR. As a result the Electoral Supervisory Commission had to use the special provision to allocate 3 more PR seats to OPR to restore the primacy of the FPTP election.

However, if only one set of vote was allowed, the result would have been different. FPR polled only 8.4 per cent of the votes in the constituencies as opposed to 10.7 per cent at the Island level. If only constituency votes were taken into consideration, FPR would not have met the 10 per cent threshold for PR with only 8.4 per cent of the votes and would not have been allocated any PR seat. The overall seat allocation would then have been 10 for OPR and 8 for MR, thus giving a majority to OPR without the application of the special provision.

Recommendation 5

It is recommended that we consider allotting the PR seats on a different basis than the current compensatory D’Hondt formula

While Model A of Sachs (allot PR seats regardless of the outcome in districts) is too little in terms of fairness (panadol to cure cancer), the current formula in Rodrigues could become too proportional in closely fought elections as in Rodrigues. The Unelected Votes Elect (UVE) sits in between the two and is more likely to generate a stable government. Simulations carried out show that it is more prone to stability than the current formula that is close to a pure proportional model.

However its impact will be limited with such a small number of PR seats.

Recommendation 6

It is recommended that we consider allocating the PR seats on a regional basis as opposed to the current Island wide formula.

It could help to create a majority. In such circumstances, we have to remove the 10 per cent legal threshold and replace it with a natural threshold based on the actual percentage of votes polled. It effectively makes it a two party system unless a third party can poll around 25 per cent of the votes. Under normal circumstances, in districts where OPR wins both FPTP seats, the PR seat will go to MR as the second party and in constituency which returns 2 MR candidates, OPR will obtain the PR seat. In case of each party winning one seat, the PR seat will go to the party with the higher share of vote in the constituency. We will obviously need six PR seats for the system to work if they are distributed on a district basis.

The above may be complicated. One way out would be to distribute 6 PR seats in two regions (as opposed to 6 on an Island basis) with each region entitled to 3 PR seats.

This is how it is done in Wales and Scotland.

Recommendation 7

It is recommended that consideration be given to changing the eligible threshold for the PR seats in view of the small number of electors in the Island.

For instance, one could lift the current threshold to 12.5 per cent of the island vote to avoid too much party fragmentation. We could also leave it at 10 per cent but base it on registered voters as opposed to votes cast as is the case in some countries. The objective is to give an advantage to the majority mode of the mixed system so that it is harder for some parties to gain seats.

However when combined with the other recommendations contained in this document, it preserves the fairness and inclusion of the formula by ensuring that one party does not control almost all seats in the Assembly.

Recommendation 8

It is recommended that we have gender parity in both FPTP and PR modes.

Women constitute slightly more than half of the population. As the Parliament of a representative democracy should primarily be a mirror of its society, it is only fair that women should have equal participation and representation.

First, the 12 FPTP candidates should be equally divided with the proviso that there must be a candidate of either gender in each of the 6 districts.

Second, on the Party list, there must be gender parity, that is an equal number of candidates of each gender. In addition there must be at least one person of a different gender out of every 2 sequential candidates.

Recommendation 9

It is recommended that the same principle of gender parity should apply for the posts of Commissioners in the Regional Assembly.

Democracy cannot truly deliver for all of its citizens and be very representative if over half of the population remains vastly underrepresented in the political process. The equitable participation of women in politics and government is essential to building and sustaining democracy. Also women have all the characteristics to be chosen as candidates and Commissioners. No more but certainly no less than men.

There must be gender parity in the posts of Commissioners in the Regional Assembly.

Recommendation 10

It is recommended that an anti-defection clause be enacted so as to prevent floor crossing.

This will enhance stability and curb blackmails and other malpractices.

While anti-defection legislation is a divisive subject, some emerging and fragile democracies have introduced such measures as defection poses a threat to their nascent democracy. The low number of seats renders the system in Rodrigues vulnerable to unscrupulous practices that could undermine the legitimacy and the stability of the system. Subject to the right safeguards that would not frustrate the freedom of individual members to express themselves according to their conscience and to prevent party leaders from abusing of their prerogatives, a member that has been elected under a party should lose his/her seat if he/she voluntarily leaves that party. The member does not lose the seat if he/she is expelled from the party. It should also be very clear who will be responsible to make the decision to declare the seat vacant. It must be an independent institution.

In countries with a large Parliament, a defection of one or two members is unlikely to create instability while in a small Assembly, it could hold the system to hostage.

Exceptionally and in view of the small size of the Assembly, I am recommending that the anti-defection clause shall apply to both FPTP and PR members.

This will significantly moderate the ardour of members to unscrupulously play with the system.

Recommendation 11

It is recommended that candidates may stand in both FPTP and PR elections.

The law does not allow an individual to stand both as a constituency candidate and to be on the PR list. Some criticisms have been leveled at the behaviour of some PR list candidates during electoral campaigns. They have an incentive to canvass for the party vote but not necessarily in favour of the candidate standing in the constituency. A narrow loss for his or her colleague increases the chance of the party list candidate to be returned on the PR list. With double candidacies, there is less incentive for such practices.

Double candidacy is also standard practice across countries with a mixed FPTP/PR formula.

One must also acknowledge the fairness of an electoral system that ensures that the top guns of the parties are returned to the Assembly. They are after all those who are responsible for framing its policies. On the other hand, the public would have considerable difficulties to understand constituency elections without the participation of these stalwarts; such contests would lose some of its “panache” and legitimacy. Also allowing some double candidacies will ensure that we avoid situations like such as in 2002 when the leader of OPR, Serge Clair, lost in the first constituency and could not be elected on the PR list of his party as double candidacy is not allowed. As a result an elected FPTP member resigned and a bye election was held to allow Serge Clair to win a seat and become the Chief Commissioner of Rodrigues. Had we allowed double candidacy, there would have been no need for such bye election and wasteful expenditures.

We should therefore strike a fair balance and allow some double candidacies. Fortunately in its wisdom, the Sachs commission recommended that some double candidacies be permitted under the mixed electoral reform proposed for Mauritius. I think it should be extended to the Rodrigues Assembly. We should consider having a cap on the number of such double candidacies to prevent abuse and to avoid the criticism that too many PR Mps are FPTP candidates who have been rejected in their constituencies. I believe a maximum of 50 per cent should be a good compromise.

The current FPTP/PR mixed system has delivered on its mandate in 2002. It also functioned extremely well in 2006 to produce a majority for the party that polled significantly more votes than its adversary. The outcome of 2012 is due to the twin reasons that OPR did not win a majority of vote and all three parties were eligible for the PR seats. The current mixed FPTP/PR voting formula can continue to serve our compatriots in Rodrigues in terms of stability, governability, fairness, inclusion and representation if some of the aforementioned recommendations are implemented.

The series of recommendations above is certainly not a set menu. Well chosen and in the right combination, it could be an “à la carte” choice to enhance the stability of the system.

This is by far the preferred option.

3.4 Option 2 : keep FPTP/PR mixed system but adopt a new formula that guarantees a fair share of seats to unsuccessful parties while ensuring that the FPTP clear winner keeps its majority.

However if some policy makers are still not convinced of a revised FPTP/PR voting formula and are bent on revising the existing formula dramatically, we should consider other alternatives as the one that is being proposed is an absolute calamity for inclusion, fairness, equity and representation of all shades of opinion in Rodrigues. We have a duty to stop it. Hence the suggestion of two second best solutions.

The main elements of the first one could run as follows:

1. to retain the 12 FPTP in 6 two-member constituencies and have either 5 additional island wide PR seats as odd numbers are more likely to produce a winner than even ones;
    
2. to provide for a gender balance of one man and one woman per constituency and a narrow zebra for the Island wide list; same for the posts of Commissioners.
    
3.  to allot the 5 PR seats on the basis of the share of votes of each party subject to certain conditions (share of votes) and depending on the circumstances;
    
4.  to maintain the 10 per cent threshold necessary for eligibility to the PR seats;
    
5. to use only one vote for both the FPTP and PR modes;
    
6. to revoke the special provision guaranteeing an overall majority to a party that has a majority of FPTP seats irrespective of its share of votes. If it is kept there must be a threshold to avoid a minority of votes being transformed into a majority of seats;

Case 1 : a party has a majority of FPTP seats (say 12-0,11-1,10-2,9-3,8-4,7-5 ) and meets the threshold of vote to qualify for a guaranteed majority after the allocation of the 5 PR seats (say 45 per cent to 50 per cent of vote)

1. the 5 seats shall be allocated in such a way that the share of seats accruing to all unsuccessful parties shall not exceed 42.5 per cent of the total, except where such is already the case after the FPTP results;
    
2. except for rounding up, an unsuccessful party may not end up with a share of seats that is higher than its share of votes except when such is the case after the FPTP results;
    
3. the 5 additional seats shall be allocated on the basis of the compensatory D’Hondt formula as is currently the case;
    
4. where there are more than two unsuccessful parties, their respective number of additional seats shall be based on their share of votes after taking into account the number of seats, if any, obtained at the FPTP mode;

Case 2 : a party has a majority of FPTP seats but does not meet the threshold of vote to qualify for a guaranteed majority after the allocation of the 5 PR seats (say 8 FPTP seats with 35 % of the votes in a three horse contest)

The current system operates as is and the 5 seats are allocated according to the compensatory D’Hondt formula. There is no special provision to award additional PR seats to manufacture an artificial majority. And as is the case in other countries with similar system, there will be post election discussion and negotiation to form a government. We can provide for the party with most seats to have the first choice at that task. This would in fact reflect the very wish of the people as no clear majority has emerged;

Case 3: it is a hung Assembly with no single party having a majority of FPTP seats (either equal number of seats by parties or no party with any majority of FPTP seats)

Again, the current system applies with the 5 PR seats allocated on the basis of the compensatory D’Hondt formula. The same process as in Case 2 will apply as it mirrors what the voters want.

3.5 Option 2 will tilt the balance in favour of stability. It will also avoid most of the anomalies and huge disproportion between vote share and seats share.

It will enhance the stability feature

A party with, say, around 47.5 per cent of the votes (to decide whether it should be fifty per cent or slightly lower) and which has a majority of seats in a three cornered contest will retain its majority after the allocation of the 5 PR seats.

It will also cure some of the worst defects of the FPTP system.

A party that polls around 40 per cent of the votes and which does not win a single FPTP seat will end up with at least 29 per cent of seats in the Assembly instead of nothing or a maximum of 13 per cent with the proposed amendments;

A party that secures fifty or more of the Island vote and which does not have a majority of FPTP seats will be rewarded with more of the additional seats to ensure that it ends up with an overall majority of seats. In the proposed amendments, the ‘loser’ will retain its lead over the vote ‘winner’.

3.6 Option 3: to keep the mixed FPTP/PR but change the formula to elect the district members

FPTP creates anomalies especially in three cornered contest as a party may win a disproportionate number of seats on less than 40 per cent of the votes. One solution would be to provide for FPTP members to be returned with at least 50 per cent of the votes. This will require either an alternative vote as it exists in Australia and a second round as in France. If no candidate goes above the 50 per cent threshold, those with the lowest vote shares are eliminated and only the first two are left in the race.

The PR seats may continue to be apportioned on the basis of the party vote as is currently the case.

It has the merit of converting the election into a two party contest in the second round or the preference voting. It will be much easier to secure an overall majority with two rather than more parties in the fray.

I am not very enthusiastic about Option 3.

I believe that Option 1 is better than Option 2 in terms of the balance between stability and fairness.

However Option 2 is much better than the alternative contained in the draft amendments.

It is also better than option 3.

Option 3 is however a significant improvement over the FPTP/Best Loser combination being proposed.

CHAPTER 4 : Concluding remarks and the way forward

There is no such thing as a perfect electoral system, even if some formulae work better than others in terms of meeting the attributes of a good electoral system. The mixed FPTP/PR formula used to return members to the Rodrigues Regional Assembly has delivered reasonably well on the twin objectives of stability and fairness as intended by its founding fathers. It worked well in 2002 and very well in 2006 while 2012 caused a problem because there was no party with a majority of vote and all three contestants were eligible for the 6 PR seats. In the light of that experience, we can improve the formula to provide some additional room for stability without reneging completely on its fairness, inclusion and representation features.

Often we cannot choose a perfect system but have to settle for one that is significantly less imperfect than other formulae. The best voting system for a country is not one that satisfies only one criterion completely, but one that provides a fair balance among the different attributes. A prudent approach is to design an electoral system that avoids serious shortcomings and flaws. The amendments being proposed are deeply flawed and constitute a backward move that will penalise diversity, restrict plurality, stifle inclusion, create unfairness and curtail political representation in a small island community.

I submit there is a compelling case to make some amendments to the current FPTP/PR voting formula to increase the chances of a majority emerging without being very unfair to political representation. We simply cannot do away completely with the PR mode as it constitutes the anchor for fairness, inclusion and representation.

This is the preferred way forward. At the end of the day, there must be a fair accurate reflection of votes in the distribution of seats in the Assembly.

However if this is not palatable to some political policy makers, there are second best solutions that can be adopted to ensure a good balance between governability and fairness instead of the disastrous alternative being contemplated. Some may look complicated but there are algorithms to deal with them in a very efficient and effective manner.

I have attempted, through research work and comparative analyses, to make an informed contribution to the debate by assessing the merits and weaknesses of the current voting formula in Rodrigues, evaluating the drawbacks of the proposed amendments, highlighting their lacunae, underlying the necessity to have a non partisan approach to electoral reforms that benefits Rodrigues and finally suggesting a pathway for reforming the system so as to provide for greater stability and effectiveness without giving up on fairness, inclusion and representation

The politics of Rodrigues is highly polarized and this shows up in election results. Any change in the electoral system to return members to the Rodrigues Regional Assembly will be seen as a zero sum game. There will be some perceived ‘winners’ and some ‘alleged’ losers. Strategic considerations and tactical calculations will dominate the agenda as each party attempts to propose what it sees as best for its short term interests. In addition, the specificity of Rodrigues, its smallness in terms of the size of the Assembly, the necessity to ensure a balance between stability and fairness of representation, have informed my analysis and the proposed changes to the voting formula .

One final consideration.

Political actors, however well intentioned, do not always have all the expert knowledge and the information to fully grasp all the consequences and ramifications of different electoral systems. Unfortunately, some may even have the tendency to believe what they know best is what is best for the country or what is best for their party is also best for the country. Worse, very often they also do not know what they do not know. This explains why countries rely on the knowledge and expertise of electoral scholars to advise them in this difficult process (UK, South Africa, Lesotho, Japan, New Zealand, etc).

As the choice of a particular electoral system will have far reaching consequences on the future political life of Rodrigues and as, once chosen, it may stay in force for a long period, we cannot afford an ill advised, half-baked or a partisan solution. We need to be alert to the unintended consequences of the proposed amendments as Rodrigues may fare worse than what it is bargaining for. It is of utmost importance for the stability and the progress of the Island that we build consensus, seek legitimacy and secure wide acceptance for a change in the electoral formula.

We also owe it to our compatriots of Rodrigues to have an informed, dispassionate and disinterested debate and to seek the expert views of external resource persons before making a decision of such importance. As it is an issue that has been comprehensively addressed in many other countries and devolved Assemblies, the independent wise people should be able to accomplish the task in a short time.

It is worth its pound of gold.

Dr Rama Sithanen,  Electoral Scholar