Roadmap for an achievable compromise to break the electoral reform deadlock

Avec le soutien de

In a series of articles to be published this week as from today, Dr Rama Sithanen explores the key points of the electoral reform issue. His last article will culminate in a national compromise among various proposals.


Part 1: Understanding the rationale, context and urgency of electoral reform

1. The enduring unresolved agenda

This set of articles provides a framework for reaching an acceptable compromise among the various stakeholders on electoral reform. We must synergise to find a better way to resolve the conflict as we seem to be going nowhere on current track. In spite of political will, many false dawns and several expert reports, committees and commissions, the country has been impotent for close to 25 years to reform the electoral system. The predicament has exacerbated with many legal setbacks against the Best Loser system in its current shape from several judicial bodies. And we are now in a very tight corner with the case brought against the State by Rezistans ek Alternativ on the mandatory declaration of a candidate’s community to stand for general elections. It will be taken on merits next week after so many postponements by Government on account of its inability to submit an acceptable solution. There has been absolutely no consultation and dialogue with other actors to find a compromise. Save for another deferment, we still do not know what will be the stand of the State.

Let us be candid. Petty short term vested interests, parochial party partisanship, unwarranted egos and unforgivable pride, tactical calculations and strategic considerations, an unconcealed lack of knowledge and understanding of how electoral systems work, a distorted view of fairness and inclusion in a diverse society, a legacy of historical prejudices on Proportional Representation, an abject refusal to learn from best international practices and a cynical disregard for the recommendations of electoral experts, all have coalesced to stifle any attempt to rise to the political and judicial challenges of reforming the voting system.

2. The putative proposals of the Ministerial committee

The Ministerial Committee set up to make recommendation on Electoral reform seems to have reached some conclusions on a package of proposals. Even if these need to be confirmed officially, it would appear that the main thrusts are as follows.

(i) keep the current 60 MPs in 20 three-member constituencies in the island of Mauritius and the single two-member district in Rodrigues;

(ii) introduce 12 seats to be returned from PR party lists;

(iii) appoint a maximum of 7 additional seats to set right any imbalance that could result from the allocation of the 12 PR seats and to provide for the replacement of the Best Loser seats;

(iv) establish a threshold of 12.5% of national vote to be eligible for the 12 and the 7 Party List seats;

(v) allot the 12 PR seats by a parallel method (Sachs Model A);

(vi) limit the distribution of the 7 seats to a maximum of 3 seats to unsuccessful parties;

(vii) provide for the 12 PR seats to be appointed from a pre-established, closed and rank-ordered Party list;

(viii) allow for the ‘up to a maximum of 7 seats’ to be chosen at the sole discretion of Party leaders on a ‘Look At’ basis.

(ix) do away with the mandatory requirement to declare one’s community to stand as candidate.

3. The political rationale for reforming the electoral system

To evaluate the efficacy of these proposals, one needs to recall the main reasons for electoral reform. There are two sets of objectives to review the current voting system. The first is the necessity to cure the imperfections of the current First Past the Post formula while the second are the various legal cases brought by Rezistans ek Alternativ against the State.

From a policy perspective, electoral reform should

(i) lower the huge disparity between votes polled and seats won by parties. In many elections, the winner often captures a very high share of seats that is unrelated to its vote share while unsuccessful parties obtain a very low share of seats compared to their vote shares. It has occurred in seven of the eleven elections since 1967 with huge discrepancies between vote shares and seat shares in 1991, 2000, 2010 and 2014 and some material vote-seat disproportionality in 1983,1987 and 2005;

(ii) prevent the phenomenon where one alliance obtains 100% of seats with around 60% of votes, thus leaving no representation at all to another party with an important vote share. It is the infamous 60-0 and has happened twice. The Labour Party did not return any candidate in 1982 on 26% of vote while the MSM did not win a single seat in 1995 with 20% of votes;

(iii) provide fair representation of women. Mauritius continues to rank among the lowest in the world in female political presence. Women are vastly and chronically discriminated both at the level of Parliament and in the Council of Ministers. They account for 50.5% of the population, yet comprise , on average since 1967, less than 10% of MPs and Ministers;

(iv) find an acceptable replacement for the Best Loser system so that our Constitution can be expunged of communal references while ensuring diversity and inclusion in political representation;

(v) cure the violation of the rights of our citizens as called for by the United Nations Human Rights Committee since 2012.

4. The Constitutional case to revisit the voting formula

The Best Loser system has been decried by many judicial institutions. The main criticisms are as follows

(a) The compulsory disclosure of a candidate’s community has been found arbitrary and discriminatory and to constitute a violation of the rights of our citizens by the United Nations Human Rights Committee on 31st August 2012. Our country was enjoined to submit an effective and enforceable remedy within 180 days so as to avoid such violation of rights in the future. Government has been incapable to find a remedy to cure such violation of rights after almost 6 years. The UN keeps reminding us of our obligations in that matter, with adverse impact on our image and reputation globally.

The Committee was referring to the fact that the eight Best Loser seats are allocated on the basis of the 1972 community census and all candidates standing for general elections must declare their community, failing which they are ineligible. It considers such practice as a violation of the rights of candidates as the census has not been updated for over 40 years (46 years now). Mauritius was adjured to choose between the two options of either updating the 1972 community census or embracing a non- communal voting system;

(b) The Best Loser system has been censured many times by the Supreme Court. In 1992, it questioned whether representation ‘is fair and adequate’ when the figures for the allocation of the 8 Best Loser seats do not reflect present reality but that of 1972. In 2005, the Supreme Court stated that:

‘it is really unfortunate that the learned Judge’s hope, as at 8 September 2000, that the defects would be remedied in the near future having regard to the project of electoral reform which was on the cards has remained a pious wish and no improvement has, up to now, been brought about.’

Thirteen years after that 2005 observation, the political class has still not found a cure for the problem identified by the Supreme Court. It still remains a pious wish.

       (c) The Privy council has chastised the Best Loser system for its use of anachronistic figures to allot the 8 seats. In 2011, it argued that

‘a system based on figures now nearly forty years old makes no sense’.

While it stated that it has no jurisdiction on the matter, it urged for the problem to be resolved through political debate and constitutional reform and warned that it will not prevent a constitutional challenge being brought against the Best Loser System in the future. It is also clear in which direction it is likely to lean;

      (d) There is also the case of Rezistans ek Alternativ against the State which is currently before the Supreme Court. Government has requested many postponements on the basis that it is working on a new electoral formula that will address the points raised on the mandatory declaration of a     candidate’s community to stand at elections. However one week before the case will be heard on merits, there is still uncertainty on the way forward.

5. Stuck in a legal quagmire

We are way behind schedule on proposing an effective and enforceable remedy to the UNHRC to prevent further violation of the rights of our citizens. Extremely late for a solution that will satisfy the various calls of the Supreme Court to reform the voting system. Behind time for a cure that will prevent a constitutional challenge being brought against the Best Loser System before the Privy Council. And very far from responding to the compelling case of Rezistans ek Alternativ in its legal action against the State.

Part 2: A critical assessment of the Ministerial Committee’s proposals

1 Hardly any proposal not already made by electoral experts since long

Many of the putative recommendations of the Ministerial Committee are not new. They have been thoroughly appraised by many independent experts from different legal, political and cultural backgrounds. From the eminent South African Constitutional Judge Sachs to the distinguished former Electoral Commissioner Tendon of India, from the respected former Mauritian Judge Ahnee to the renowned French Constitutionalist Carcassonne, from the reputable British Professor Bogdanor of Oxford University to the influential Spanish Professor of political science Vilanova. And from the Select Committee of the National Assembly of 2002 to the White Paper of 2014 that consulted many UK and other electoral experts. However, it is the first time that a Ministerial Committee is considering this vexed subject on its own, without the assistance of independent electoral experts or a technical committee to advice on thorny issues. Also there has been no consultation and dialogue with other political parties, civil society and academia.

2 Most proposals flatly rejected by ALL electoral experts

The proposals of the Ministerial Committee are completely at odds with the recommendations of all these experts in electoral systems. Whether this is due to a lack of domain knowledge, the absence of any technical advice and insights from electoral specialists, an attempt to embrace partisanship or a trial balloon, is anybody’s guess. Unsurprisingly all stakeholders who have commented on these proposals have bluntly dismissed them. To be fair, it is not difficult to understand this unanimous rebuff.

3 Too few seats to correct the four deficiencies of the FPTP formula

Most experts found that an absolute minimum of 20 proportional representation (PR) seats are necessary to meet four of the five objectives of the reform. The 12 seats suggested by the committee are far too few to be fair to parties, to women and to subsume the 8 Best Loser (BL) seats. It should be pointed out that Sachs recommended 30 PR seats while the initial proposal of the Mouvement militant mauricien (MMM) was 28 PR seats and the Labour Party (LP) advocated 17 such seats;

The “up to a maximum of 7 seats” allotted on a restricted basis will be very inadequate to replace the 8 BL seats in order to ensure diversity of representation and inclusion. The more so that in some elections, such as in 1987, no such seat will be distributed as the initial difference in seats between winner and loser may not be affected after the distribution of the 12 PR seats. If no or very few of the 7 seats are distributed, it would be an absolute calamity in terms of diversity of representation and inclusion for some important components of our population. Sachs had proposed 30 seats in one list. The LP in its discussion with the MMM wanted one party list of 20 Members of Parliament (MP), while the MMM opted for two separate party lists of 28 MPs. They eventually compromised on two party lists of 14 and 6 PR seats respectively.

4 The parallel mode is very weak to ensure fairness

The parallel method of allocating the party list seats (Sachs A) has been totally rejected by all electoral experts since 2002. Sachs found even 30 PR seats distributed by the parallel mode too weak to address the unfairness of the system and ranked it last among the five alternatives it considered. Professor Chu from Singapore called that mode “panadol to cure cancer” as its effects on the inequity of the First Past The Post (FPTP) is only symbolical. If 30 seats with the parallel mode was turned down, one can imagine how they would characterise 12 seats with that same allocation system. Simply disastrous. Collendavelloo also spurned the parallel mode in the Select Committee of 2002 precisely because it does not “cure the disease”.

The reason is very simple. Instead of narrowing the disparity in seats between winner and loser, the parallel mode in fact gives more PR seats to the party that has already benefitted disproportionately from the FPTP system. Sachs recommended the compensatory mode to lower the discrepancy between winner and loser. The MMM agreed with the Sachs proposal while the LP felt that the compensatory mode could be too proportional in some instances, thus making it difficult to produce a majority to govern the country with stability. It instead suggested an allocation method based on the principle of “every vote counts”.

5 The oddity of a 12.5% eligible threshold

I am very surprised at the alleged hurdle of 12.5% of national vote to be eligible for the 12 PR seats. If true, it is far too high and very different from proposals made by the Mouvement socialiste militant (MSM) in the past. I do not think it exists in any country as it is very prohibitive. It cannot also be too low. The choice is a difficult one between being fair to small parties and the need to prevent a fragmentation of the political system based on very divisive criteria in a plural and diverse society. After an exhaustive evaluation of these two critical considerations, Sachs had absolutely no hesitation to recommend a 10% threshold.

6 Undemocratic for leaders to choose up to 7 MPs on a “look at” basis

There is no country with a dose of PR that has two party lists, one chosen by the electorate on the basis of a closed and rank-based list and one appointed at the sole discretion of party leaders after polling. In all countries, a well-balanced single party list ensures fair representation to all components of society. Sachs and Carcassonne had absolutely no hesitation to go for a single closed and rank-based party list. During the 2014 discussion, the LP argued that diversity could be achieved with one list only, as contained in the government White Paper. Nevertheless, a compromise of a first list of 14 MPs chosen by the electorate and a second list of six MPs appointed by party leaders was found between the LP and the MMM. However, it confers too much power and influence to party leaders to use their discretion to appoint MPs. It is highly undemocratic and it removes the absolute right of the electorate to choose their MPs.

7 Clarity required on some issues

It is unclear how the Ministerial Committee proposes to appoint the “up to seven MPs” which apparently have the twin objectives of restoring the difference in seats between the winner and the unsuccessful parties and to replace the 8 BL seats. Especially as there are elections when no such MPs will be chosen. We need certainty on the restriction of 3 MPs out of the 7 being imposed on unsuccessful parties. How will such restriction be operationalised?

What will be the basis on which party leaders will appoint the “up to seven MPs”? Will community belonging constitute a covert factor in such choice? Will different shades within one single community be acknowledged? Will there be guidelines from a body to help leaders make their choice? Or will it be left at the absolute discretion of party leaders?

What are the recommendations of the committee for women’s participation and inclusion both at the level of the National Assembly and at Cabinet? Are they at least one third of all MPs or are they materially different?

Will the reform be introduced at the next general elections? Surely, nobody will accept a reform enacted today that will be applied only in 2025. Anyway, it will not address the legal issues before the Supreme Court.

8 Back to the drawing board again

As most, if not all, of the proposals of the Ministerial Committee have been comprehensively discarded by all electoral experts, it is no surprise that many parties have turned down these recommendations. The Parti mauricien social-démocrate, MMM, LP and Mouvement patriotique have all dismissed the content of the reform as “inadequate, insufficient, deeply flawed and unacceptable”. The MSM/Muvman Liberater alliance does not have the requisite majority to change the Constitution to force through its reform agenda. We are therefore back to the drawing board and it is unclear what will be the stand of the State when the case against it is called before the Supreme Court on the 24th of May. We must therefore craft a compromise that stands a chance of securing a three-quarter majority in the National Assembly.


Part 3: Picking optimally from three sets of proposals to reach a consensus

1. Choosing from Ministerial Committee, LP and the MMM and Sachs Commission

There is urgency to move out of the tight corner where we have locked ourselves up for close to 25 years. We should attempt to find a compromise from among the different recommendations that are on the table. I shall consider three such proposals.

(i) First are those of the Ministerial Committee;

(ii) Second is the agreement reached between the Labour Party and the MMM after protracted and detailed discussion and negotiations in 2014. It did not represent their initial position but was the outcome of a compromise to find a balancing act. As it should be on such subject;

(iii) Third are the recommendations of the Sachs/Tendon/Ahnee Commission of 2002 which remains a reference

The Table below shows the three sets of proposals which will be used to craft a compromise that could constitute the basis of negotiation among the various political stakeholders to finalise the architecture of a Bill likely to command the required majority of three quarters in the National Assembly.

2. 62 or 63 FPTP seats

The Ministerial Committee is proposing to retain the current 62 First Past The Post (FPTP) seats, made up of 60 seats in 20 three-member constituencies in the Island of Mauritius and a single twomember riding in Rodrigues. Sachs also kept the same number of FPTP seats. However one has to acknowledge three facts that play in favour of the same three-member constituency in Rodrigues, thus increasing the number of FPTP seats by one from 62 to 63. First, Rodrigues is not anymore the constituency with the lowest number of electors. Second, parties in Rodrigues are unlikely to benefit from the proportional representation (PR) seats as they will not reach the share of national vote to be eligible. Third, an additional MP will greatly facilitate the inclusion of women. One way of making the reform participative and inclusive is to have a threemember constituency in Rodrigues. It is a small concession.

3. The number of PR Party List seats

As concluded by all electoral experts, 12 PR seats are far too low to achieve the main objectives of reform. The Ministerial Committee could agree to fold the two lists of 12 and ‘up to a maximum of 7 seats’ into one list of 19 PR seats. This should not pose a problem for the Committee as there would be some elections when the 7 seats would be distributed. While being a fair compromise, it is still far from the 30 seats of Sachs and the 28 seats of the MMM.

If there were too many PR seats, the stability of the system would be eroded 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 were too few PR seats, it might be hard to lower the disproportionality of the FPTP mode, to ensure gender fairness and to subsume the Best Loser System. According to Sachs, fewer than 30 PR seats would be inadequate to remedy the anomalies of FPTP. Detailed simulations show that 30 or 28 PR seats are too high while 12 seats are far too low. The optimal number of PR MPs would be around 20 to 25 depending on a host of factors ranging from the mathematical formula used to allot PR seats and the threshold of votes required to be eligible for PR seats, amongst others. If we accept the bare minimum as 20, a difference of one such seat between the Committee’s proposal and the LP and MMM agreed number in 2014 should not be a deal breaker. While I would suggest 20 PR seats, we could live with the 19 seats of the Ministerial Committee.

4. One single party list is democratic, accountable, transparent and best practice

The basis to have two party lists, one elected by the people and the other chosen by Party leaders after the elections, is highly controversial in a democracy as it confers too much power, influence and discretion to one single individual. It purports to leave the choice to the judgement of political leaders to correct for any ‘community’ underrepresentation that could occur after the election of the 63 FPTP MPs and the 20 Party List MPs. Experience across the world shows that a well-balanced single party list should deliver on diversity of representation and inclusion. This is best practice in many plural and diverse countries, including South Africa, Belgium and Northern Ireland.

I completely agree with Sachs, Tendon, Ahnee, Carcassonne, Bogdanor and Vilanova who are against the selection of some MPs on a ‘Look At’ basis by party leaders after the elections. I am unaware of any democratic country where after the results of a non-ethnic poll are declared, ethnic considerations are introduced by political leaders to designate some representation of communities considered to be under-represented. It also brings back the very bad memories of the old colonial days with the prerogative of the then Governor to appoint some MPs on a ‘Look At’ basis after the elections. The system was absolutely undemocratic, extremely subjective, highly discretionary, hugely non-transparent, terribly unaccountable and largely open to very corrupt practices. One of the best constitutionalists who visited Mauritius (Professor de Smith) strongly recommended its discontinuation for all these reasons.

Parties should constitute their list both in terms of numbers and rank to broadly reflect the diversity and plurality of the country and cut across religious and ethnic divides. And the single party list must be established and submitted prior to elections to the Electoral Commissioner’s Office.

5. The threshold for eligibility to PR seats

The choice of a given threshold depends on the level of political fragmentation we are willing to tolerate, especially with respect to communal, religious, racial and casteist exacerbations.We need a minimum threshold of national vote for parties to be eligible for the 20 PR seats so as to preserve social harmony and national cohesion and ensure stability. It is a difficult balancing act between the genuine democratic rights of small parties that have a key role in our political landscape and the high risks of an atomisation of the system. 12.5% is extremely high and will prevent legitimate views to be expressed in Parliament. However it should not be too low as it could encourage a balkanisation of the country. Similar to Sachs, I prefer to be safe than sorry and therefore recommend a threshold of 10% of national vote to be entitled to PR seats. Sachs, Tendon and Ahnee who were very staunch supporters of small parties, saw the risks of fragmenting the political system and recommended a 10% hurdle. Other Commissions in the past have also proposed a 10% threshold.

6. Meeting the requirement of the UNHRC, the Supreme Court and the Privy council

As the Best Loser system would be replaced by a new dispensation, there would be no need for our citizens to disclose their community belonging to stand as candidates at General Elections. Equally there would be no necessity to use the population census of 1972 to award the 8 Best Loser seats. As a matter of fact, we would stop categorising our compatriots into four ethnic/religious silos on the basis of their ‘way of life’. Rezistans ek Alternativ should be commended for this relentless fight to deethnicise our Constitution and our electoral system.

Part 4: Allotting PR seats for stability, fairness and retention of FPTP absolute majority

1. A crucial decider of the stability-inclusion nexus

One of the key drivers of the balance between government stability and party fairness is the mode of distributing the PR seats. There are essentially two methods of allocating PR seats, the parallel mode (Sachs Model A) as proposed by the Ministerial Committee and the compensatory formula as recommended by the Sachs Commission (Sachs Model C). In view of the weaknesses of both methods, I have suggested a new algorithm which represents a compromise between these two extreme modes. It is based on the principle of “every vote counts”.

While many grasp the difference between FPTP and PR systems, they do not always make the fundamental distinction between a parallel PR and a compensatory PR, when used in tandem with FPTP. For instance, 20 PR seats apportioned through the compensatory method is much fairer to parties than 40 PR seats using the parallel mode. We must strike a good balance between fairness to parties in terms of the vote to seat ratio and the stability and governability of the system. If the formula were too fair such as in Rodrigues (Sachs C), it could sacrifice stability and effectiveness in closely contested elections as it tilts too much in favour of parties that have been penalised by the outcome of the FPTP elections. If it were too timid in attenuating the unfairness of the FPTP, the defect of huge distortion between votes polled and seats obtained would remain as it favours the winner of the election too much.

2. The parallel mode is very inadequate to cure the unfairness of FPTP

All experts have dismissed the parallel formula as it is very insignificant in curing the deficiencies of the FPTP system. It leaves a huge distortion between vote shares and seat shares. The presence of the Opposition becomes symbolical only and it would not be able to play a meaningful role in Parliament. This is the main reason why it was forcefully rejected by the Sachs Commission and the Select Committee chaired by Collendavelloo and not even considered by Carcassonne.

As aptly put by Dr Chu, “it highlights the impotence of parallel PR to tone down the excess of FPTP attribution… Worst still, it has the perverse effect of increasing the lead acquired via FPTP seat”.

However, the compensatory mode of allocating PR seats as advocated by Model C of Sachs could convert the electoral system into a full proportional one in some cases and thus make it more difficult to have stable government.

The Table below illustrates the major difference between the parallel and the compensatory method of distributing PR seats in a mixed FPTP/ PR electoral model. The election of 1995 is used as an example.

When the parallel mode is Panadol to cure cancer

The parallel mode, which disregards the fact that the MSM/ RMM did not receive a single seat for 19.7 % of the votes in FPTP elections, would allocate 5 PR seats to the second party and 15 PR seats to the LP/MMM alliance, based on their respective share of votes. It is heavily biased in favour of the party that has already benefitted massively from the FPTP formula with 100 % of seats on 65.2 % of votes. The final outcome is 75 seats for the LP/MMM and only 5 seats for the MSM/RMM. It is highly disproportional with the MSM/RMM attracting only 6 % of seats for 19.7 % of votes.

The compensatory method recognises the unfairness of the FPTP elections and rewards the MSM/RMM for its share of votes. It would receive 18 of the 20 PR seats as it was severely penalised in the constituencies. The final result is 62 seats for LP/MMM and 18 seats for the MSM/RMM. The second party obtains 22 % of seats for around 20 % of votes (the difference of 2 % is due to the threshold of 10 %). In 1995 the compensatory mode would have been fair and it also leaves a huge working majority of 40 seats for the winning team. The parallel formula is very insignificant in curing the unfairness of the system with only 5 out of 82 seats to the MSM/RMM.

3. The compensatory mode could imperil the stability of the system

However, the compensatory mode of allocating PR seats could convert the electoral system into a full proportional one in some cases and thus make it more difficult to have a stable government. This is shown below with the example of the 1987 elections.

When the compensatory mode becomes a threat to stability

With Sachs Model C, most, if not all of the PR seats, would accrue to the party that has suffered most from the unfairness of the FPTP. It happened in 1987 when the MMM received only 34.3 % of seats with 48.12 % of the vote while the MSM/LP/PMSD won 62.9 % of seats with only 49.86 % of the vote. In view of the huge disproportionality of the FPTP outcome, the PR seats would have been awarded mostly to the MMM (18 PR seats against 2 for the winning party). The consequence is the significant reduction of the MSM/LP/PMSD majority from 18 seats after the FPTP results to only a majority of only 2 seats after the PR seats allocation .Worse, if the two MPs from Rodrigues are not counted as forming part of the winning alliance, there is a deadlock with 41 seats for the Government and a combined 41 MPs against it. This is the main reason why the MSM and the LP are against the compensatory mode of Sachs C. It could transform a large seat majority into either a very thin lead or a hung Parliament. The outcome of elections in Rodrigues has also worked against using the compensatory mode.

4.The “every vote counts” compromise

A simple example will drive home this point.

In 2005, P. Jugnauth polled 48.297 % of the vote and did not win a seat while S.Moutia became MP with slightly more vote at 48.415 %. Jugnauth’s 48.297 % vote counted for nothing because of the FPTP mode. The same occurred between M. Hanoomanjee and E. Jhuboo in 2014 and between V. Bunwaree and I. Collendavelloo in 2000. Very few votes separated them and yet the difference is huge between being ‘in’ and ‘out’. The vote for one counts very highly while that of the other has a zero value. An electoral system which counts ALL votes can be considered desirable on grounds of fairness.

5. How it delivers on both stability and fairness

The table below shows how the “every vote counts” mode delivers a working majority even in closely contested elections. It is an improvement on the compensatory method while being fairer than the parallel formula.

When “every vote counts” strikes a balance between parallel and compensatory modes

While the compensatory mode would give 18 PR seats to the MMM and only 2 PR seats to the winning team in 1987 , the “every vote counts” method allocates 14 such seats to the MMM and 6 seats to the MSM/LP/PMSD . It is much fairer than the parallel mode as it compensates the MMM better with 14 seats instead of 10; however it is more stable as it delivers a working majority of 10 for the MSM/LP/PMSD, which is not the case with the compensatory formula. The formula has the advantage of rewarding the overall winner also in constituencies where it has failed to win seats.

6. Responding to MSM’s concern on retaining the FPTP majority

The “every vote counts” formula is likely to work in almost all cases to ensure there is no reversal of FPTP majority after the allocation of the 20 PR seats. We could even improve this probability by using the percentages of vote instead of the absolute vote of each party in the 21 constituencies to allot the PR seats.

If this did not break the deadlock, there could be one further assurance.

In view of the differences in the number of voters in the 21 constituencies, there could be some rare cases of closely contested elections where a FPTP majority reversal is possible. For instance if an alliance took 32 FPTP seats (out of 62) by narrowly winning all seats in 10 of the largest constituencies and 2 seats in another riding, and lost by a wide margin in the 9 smallest constituencies, its majority of 2 seats could disappear as the other party would take more PR seats because of its higher share of unutilised votes. If this were considered unsatisfactory, we could introduce a rider to the formula as follows:

(i) a party that wins an absolute majority with 32 to 35 FPTP seats shall not have a lower margin of victory in absolute seats after the distribution of the 20 PR seats. It means that a FPTP outcome of 32 seats for Party A to 30 seats for other Parties will deliver at least 42 seats to Party A after the allotment of the 20 PR seats against a maximum of 40 seats to other Parties; and

(ii) a party that wins an absolute majority with at least 36 FPTP seats shall not end up with a lower margin of victory than 10 seats after the distribution of the 20 PR seats. It ensures that a FPTP outcome of 39 seats for Party A and 23 seats for other Parties will give at least 46 seats to Party A and 36 seats to other Parties.

Such safeguards would not be necessary in most cases as the system is likely to be self regulatory. However as compromises are the order of the day, it has the merit of allaying the fears of some political actors strongly opposed to a reversal of a FPTPmajority after the distribution of the 20 PR seats.

Part 5: The safeguards of the London Agreement and the Constitution on ‘fair and adequate’ representation

1. The Constitutional dilemma: either update 1972 census or endorse a non-community based voting system

Let me dispel a popular misgiving. Neither the UNHRC nor the Supreme Court, nor the Privy Council has condemned the Best Loser system (BLS) for its ethnic/religious classification of the population. They have simply criticised the fact that the eight BL seats are being allocated today.

‘without the corresponding updated figures of the community affiliation of the population’

as posited by the UNHRC in 2012. The Supreme Court and the Privy Council have also castigated it because of the outdated 1972 statistics used.

It is therefore not a surprise that to prevent further violation of the rights of our citizens, the UNHRC enjoined our country to choose between

updating the 1972 community census or considering whether the community based electoral system is still necessary’.

Strictly speaking, an update of the 1972 population census would be acceptable to the UNHRC, the Supreme court and the Privy council to deal with the legal case of Rezistans ek Alternativ. However it would leave the other aspects of reform totally unresolved

Similar to the polarising debate in the 50’s and 60’s, political parties are sharply opposed on which option to choose to satisfy the UNHRC. The PMSD and the MSM are in favour of the first alternative of having a new population census as they believe the BLS is the best safeguard for the ‘fair and adequate representation’ of the four communities. Some stakeholders such as Shakeel and Yousuf Mohamed and some community leaders embrace a more subtle stand by arguing that the BLS should be replaced only if we find a system that offers similar guarantees to ‘minorities’ in terms of Parliamentary representation. All other political parties and probably the majority of Mauritians are in favour of a non-com- munity based voting system. However, as well articulated by Sachs, the huge challenge facing those supporting a non-communal system is to produce a reform that affords safeguards to communities that have benefitted from the BL system. Some of the best political, legal and constitutional brains had considerable difficulties to find such a voting formula in the 50’s and the 60’s. It is not different today.

2. The guarantees of the London Agreement, Trustram-Eve and Banwell

This deep division on the electoral system is indeed a very old one. It underpinned the discussion in 1957 for the first election held with universal adult suffrage in 1959 and the debate in 1965 preceding our Independence in 1968. The London Agreement of 1957 attempted to reconcile the sharp disagreement between supporters of strong and stable government and advocates of broad social, political, and ethnic representation by proposing a seemingly contradictory compromise. It agreed on a formula that

‘should facilitate the development of voting on grounds of political principle and party rather than on race or religion’ with the safeguard that ‘any system of voting should provide an adequate opportunity for all the main sections of opinion in Mauritius to elect their representatives to the Legislative Council in numbers broadly corresponding to their own weight in the community’.

That twin objective became the basis for the Trustram-Eve commission that proposed the voting system of 1959. For the avoidance of doubt, the Commission made it clear that ‘all the main sections of opinion’ stated in the second principle of the London agreement cannot mean anything other than ‘each main section of the population’; it also agreed that ‘in numbers corresponding to their own weight’ is equivalent to ‘corresponding to its own number in the community as a whole’.

The electoral system chosen was 40 First Past The Post (FPTP) single-member constituencies combined with the right of the Governor to appoint up to a maximum of 12 members after the elections to correct for ethnic underrepresentation. In two general elections (1959 and 1963), this ethnic proportionality was achieved by the Governor using his power to secure broad communal representation that could not be attained by the elections themselves.

Dissatisfaction with both the single-member constituencies and the Governor’s appointment led to the Banwell commission being set up in 1965 to consider the voting system for the Independence elections. Three of its terms of reference were the introduction of multi-member constituencies, the discontinuation of nomination by the Governor and importantly

 ‘the system should give the main sections of the population an opportunity of securing fair representation of their interest, if necessary by the reservation of seats’.

After bitter divergence on how to meet the above requirements, the country ended up with the current 62 FPTP seats in 21 constituencies and 8 Best Loser seats to remedy for community underrepresentation. We luckily avoided reserved seats and separate electoral rolls as they exist in Fiji!

3. Constitutional entrenchment of such guarantees to ‘each community’

With the 8 BL seats, our unique FPTP formula in fact transforms the electoral outcome into a proportional representation system based on a mix of religion (Hindu and Muslim) and ethnicity (General Population and Sino-Mauritian). This is explicitly stated in our Constitution. Clause 5 (1) of the First Schedule of the Constitution speaks of the need

‘to ensure a fair and adequate representation of each community’.

The Constitution goes on to explain in fine details how such fair and adequate representation should mathematically be achieved by comparing the share of each of the four communities in the population census to the share of the 62 FPTP elected representatives. The identified underrepresented communities are then compensated with Best Loser seats until either their census ratio is reached or there are not enough seats to make it a full proportional system. For instance, after the distribution of the 8 BL seats in 1967, the results became a pure ethnic/ religious PR system in terms of the ratio of seats to the population of the four communities. It was also a very close communal PR in 1983, 2000 and 2010.

4. The apprehensions of losing such safeguards

As it is a very cleaved debate, many are fearful of losing the guarantees of the London agreement and the safeguards of the Constitution as a result of the replacement of the Best Loser system. There is thus a compelling case to reassure them as no reform will work if it is not seen as fair, adequate and inclusive. For instance, in spite of assurances to the contrary, few specific groups that were arbitrarily absorbed into the four-way communal classification in 1967 have seen a chronic erosion of both their Parliamentary and Ministerial representation as they were denied separate eligibility for the BL seats.

Such disquiet has led some stakeholders to strongly argue for the maintenance of the BLS and an update of the 1972 population census to meet the requirements of the UNHRC, the Privy Council and the Supreme Court, which is an option even if many are robustly opposed to it. Few have gone as far as asking for a review of all constituency boundaries for a better balance among the communities while others seek comfort in the appointment of few MPs on a ‘Look At’ basis . After fifty years of Independence, most people are probably in favour of a non-communal voting system. However, the formula should rise to the challenge of affording ‘fair and adequate representation’ to all the colours of the Mauritian rainbow.

5. Beware of the spectre of a new census as a default solution

In the absence of a non-community based electoral system that provides ‘fair and adequate representation’ as contained in our Constitution, the country may have no choice than to choose the other cure suggested by the UNHRC. In addition to stoking the feeling of communalism and inhibiting nation building and the several patent defects of the BLS, its retention will necessitate a new population census that will categorise our compatriots probably not into four but many more pigeon holes as the four communities were arbitrarily imposed by the Banwell Commission. At least four other sections of the population were refused eligibility to BL seats. Those who strongly believe they have been short changed by this exclusion will legitimately fight for separate identity to access the BL seats. It would be tantamount to opening a can of worms that could lead to a dreadful division of our country.

6. The need for sharing to ensure an acceptable replacement of the BL system

There is therefore an absolute necessity for safeguards in the mixed FPTP/proportional representation (PR) formula to allay the fears of some people. It is plain that some PR algorithms have a better chance of achieving ‘fair and adequate’ representation’ than others. It is not an easy task as all communities would likely benefit from the PR seats while the current BL seats accrue primarily to two communities only! It very much depends on the construction and safeguards of the PR mode . Otherwise the new formula will not and cannot replace the assurance given by the BL system. For instance, the ‘up to seven seats’ of the Ministerial Committee would not at all provide ‘fair and adequate representation’ to all sections of our population as contained in the Constitution.


Part 6: Evolve a replacement of the BLS that produces ‘fair and adequate’ representation

Political forces that are against the classification of our nation into many ethnic and religious groups after 50 years of Independence should make some compromises for the adoption of a non community-based voting system recommended by the UNHCR. I am suggesting some specially tailored algorithms, few institutional safeguards and confidence-building measures that work in other countries to ensure broad based representation and inclusion.

1. Keep the three member constituencies with a mandatory three votes per elector

It is recognised that the three-member constituencies and the three compulsory votes in our FPTP system are good in terms of encouraging a balanced slate of candidates. It emphasises diversity, encourages adequate representation and gives a fair chance to many groups to gain political representation. This would be very difficult with single member ridings.

2. Retain unequal constituencies until the new system is embedded

To encourage diversity and plurality, especially for groups which are dispersed across the country, our FPTP system accepts huge variation in the voter density of the 20 districts. It is a fact that some sections of our population benefit from gerrymandering while others are advantaged by malapportionment. Some political leaders want to revisit the boundaries of the 20 constituencies while others would like to have much lower disparity in voters in the 20 electoral districts. Those would be extremely divisive exercises. I would recommend that we do not disturb too much the number of electors in some constituencies for the sake of diversity of representation and inclusion. This is the price to pay in a FPTP electoral system to achieve ‘fair and adequate representation’. We should keep the size and composition of constituencies mostly as they are until reform has settled in. It would avoid an untimely dispute on electoral boundaries. However, it should not be too difficult to accommodate the periodic review that is required by the Constitution.

3. Choose 20 PR seats from a closed and rankbased Party list

Both the number of seats and the composition of Party lists are important for diversity and inclusion. We need 20 PR seats elected from a closed and rank-based Party list. This is similar to what was proposed by Carcassonne and Sachs. While the list would be an inclusive one, the closed and rank-based nature is one way of providing assurance to those who feel concerned by the replacement of the BLS. The order of candidates on the closed Party list cannot be changed by voters.

4. Allow parties to field double candidacies

In many countries with mixed electoral systems, there is provision for double candidacies where some candidates are fielded in FPTP constituencies and are also included on the party list. The basis for such flexibility is so obvious in our electoral system. It acts as an alternative to the BLS by giving a second chance to some candidates who are not elected while their running mates are returned. Split tickets happened in a very high number of 8 constituencies in 2014. In some cases, it did not matter as it did not affect diversity and inclusion while in others it mattered significantly and the BLS allowed for a correction. The split tickets in constituencies 2 and 3 would have no impact in the absence of the BLS while it would have been hugely important in constituencies 1, 12 and 20. Without double candidacies, Wong, Henry, Lepoigneur and Quirin would not have been MPs. And it is not obvious who would have replaced them on the Party list!

With double candidacies, those elected on the FPTP voting mode will have their name struck from the party list and other candidates will be chosen as PR MPs based on their rank order. It is a selfregulating mechanism and gives a good chance even to lowly ranked candidates to be elected from the party list. And it works well in many countries with mixed system. Otherwise, an unelected candidate in a mixed constituency could be replaced by someone with a different profile on the party list. It would go against the principle of replacing the BL system.

By the way, the split tickets in the other three constituencies (13,18 and 19) shows in a very crude manner how the exclusion of one of the five communities from the BLS in 1967 affects broad based representation. Had that group not been arbitrarily excluded, Chetty from constituency 18 would have been the second best loser and Maistry from 19 the fourth one while Nagalingum would have been returned as the 8th BL member. Because that group was acutely underrepresented with only two elected representatives (Sawmynaden and Baloomoody) from the 62 FPTP MPs. Three BL seats would have increased that group’s representation by 150%.

5. Provide for some candidates to be placed on the same rank on an ‘either or’ basis

Some countries allow two candidates to be placed on the same rank on the PR list on an either/or basis so as to give a fillip to diversity and inclusion. It has the advantage of giving a higher ranking to candidates from some groups – minorities and women – who may have difficulties to be elected. We could allow a limited number of ‘same-ranked’ candidates on the party list.

6. Give some limited flexibility in the choice of few PR seats

It is a measure of flexibility to appoint few MPs to redress a blatant lack of diversity and inclusion. For instance, there could be the option for the last four or five of the 20 PR seats on a single party list to be allocated on a closed but not necessarily a rank-ordered basis. Those familiar with the algorithm of electoral system will fully appreciate how this will impinge on the diversity of Parliament. However, the responsibility for such appointment should be entrusted to an independent body such as the Electoral Supervisory Commission, albeit after consultation with parties which have won such seats.

7. Combine fair and adequate representation with merit and expertise

This is probably one of the best ways to guarantee fair and adequate representation at the highest level in deeply multi-cultural societies. While no democratic country provides for party leaders to appoint MPs after an election, there are so many that allow the President or the Prime Minister to choose Ministers from outside the Parliamentary caucus to ensure diversity and inclusion. At the same time, it allows for Ministers to be selected from a broader reservoir of talents, experience, skills and expertise. Mauritius falls into a very small category of countries that severely restrict its choice of Ministers from an extremely narrow pool of candidates. Almost all countries recruit their ministers from a much wider pool of skills and competence. France and Italy often appoint members from ‘minority groups’ as Ministers to display diversity and inclusiveness. Kenya changed its constitution for all Ministers to be appointed from outside Parliament and uses it also to have a balance among the different ethnic groups of the country. Morocco, Rwanda, Ethiopia, and all French Speaking African countries appoint Ministers from outside Parliament for their expertise, experience and skills while promoting diversity and inclusion. The system has the double advantage of diversity and inclusion while picking talent and expertise for the various posts. Already, the Attorney- General and the Speaker in Mauritius may be appointed from outside Parliament. Our Best Loser System also allows unsuccessful candidates to be returned to Parliament and to become Minister. While we do not have to follow France or Kenya, one compromise would be to give the flexibility to the PM for the appointment of a limited number of Ministers from civil society, NGO and other individuals based on their merit, expertise and experience in their specific fields. And to use this mechanism to also provide for fair and adequate representation whenever necessary.

8. Field a good mix of candidates in the 20 constituencies

The FPTP will remain the core of the voting system with 63 seats out of 83 (76 %). So, it is important for political parties to assume their responsibility by having an equitable socio-demographic mix of candidates in the 20 constituencies so that the slate is a diverse one in terms of ‘fair and adequate representation’ of all the colours of the rainbow. This can be achieved in terms of the choice in each constituency, a national balance among the 60 candidates and the composition and rank of the party lists.

9. Grant women their rights and have a Government that looks more like the country it serves

The country simply cannot continue to deny women their legitimate rights in political representation. And also their vital contribution to the country. Time is up! We should adopt a very simple set of measures. First, we should provide for the list of candidates in each constituency to comprise no more than 2 persons of the same gender. Second, on the Party list, we should ensure that neither gender represents less than 33% of candidates. Third, there should be at least one person of a different gender out of every 3 sequential party list candidates. Fourth, neither gender should constitute less than 33% of Ministers, PPS and other Parliamentary posts . This is also ‘fair and adequate representation’ to 50,5 % of our population! When Justin Trudeau brought more women and minorities in his Cabinet, he stated that ‘we are in 2016 and the government should look more like the country it serves’. We should do the same!

10. Concluding remarks 

As there is no perfect electoral system, the roadmap proposed in these articles is certainly not devoid of flaws. However, it has the features to take the country forward in many ways and can constitute a strong basis to engage in an informed discussion with all stakeholders to find the architecture of a Bill that could command the requisite majority in the House. It deethnicises our Constitution and our electoral system while being fair to political parties and equitable to women. It also provides the assurance to achieve the original objectives of the BLS in terms of fair and adequate representation of all sections of the population.

It is now time for informed judgement, lucidity and a degree of sharing and tolerance to take over and show the way ahead. We can provide an effective and enforceable noncommunal remedy to the UNHRC so that the fundamental rights of our citizens are upheld, while making it possible for diversity, inclusion and broad based political representation to bloom without the opprobrium of constitutional community classification. This reform package could be seen as a compromise to ensure ‘fair and adequate representation’ as spelt out in our Constitution. It also avoids a descent into hell with a new population census which is the alternative cure proposed by the UNHRC. With all its risks.



{{#if summary}}


{{/if}} {{#if image}}
{{/if}} {{{body}}}
Lexpress deals logo

Retrouvez les meilleurs deals de vos enseignes préférées.

Lexpress deals brochure
Rejoignez la conversation en laissant un commentaire ci-dessous.

Ailleurs sur

Les plus...

  • Lus
  • Commentés
Suivez le meilleur de
l'actualité à l'île Maurice

Inscrivez-vous à la newsletter pour le meilleur de l'info

Pour prévenir tout abus, nous exigeons que vous confirmiez votre abonnement

Plus tardNe plus afficher