According to many, the Best Loser system is most undemocratic and is incompatible with the spirit and the letter of the Constitution. They argue that it is an aberration, at a time when real attempts are being made at nation-building, to institutionalise communalism in our Constitution. It is against republican values and ?mauritianhood? to have ethnic considerations written in the constitution for the designation of Best Losers and that this is tantamount to allowing the cancer of communalism to vitiate the very foundation of our democratic framework.The BLS ethnicises the electoral system, legitimises communalism and inhibits nation building.
As Sachs put it: ?the best approach would be to devise alternative and credible means of providing reassurance to the community or communities most supportive of the Best Loser system ? it should be possible to achieve the original objectives of the BLS in a manner that is consistent with contemporary political reality?
He argued that it is possible to introduce a dose of PR in the current electoral system and ensure that the party lists reflect the ethnic diversity of the country. For instance without the Best Loser system, Razack Mohamed, a leading member of the Muslim community, would have entered Parliament if there were a dose of Proportional Representation in 1967. In 1983 Bérenger would have been the first person on the PR list of the MMM as he was the leader and the Prime Ministerial candidate. In 1987 the same Berenger would have been returned through a PR list while he could not enter Parliament with the Best Loser system, in spite of polling almost 49 per cent of the votes. It is possible to achieve broadly representative results with some PR list that reflects the diversity and plurality of Mauritian society. With the right PR list, Koenig, the leader of the PMSD in 1967, would have been chosen, Mohamed in 1976, Ramgoolam in 1982, Bérenger in 1987 and Jugnauth in 1995. Yet none of these political stalwarts won a seat in Parliament with the Best Loser system.
Unfair and discriminatory
The arbitrary choice of only four eligible communities is discriminatory as it does not reflect the sociological realities of present day Mauritius. Certainly not in terms of our policies on languages,cultural diversity,religious subsidies,public holidays and even alignment of candidates at elections ,amongst others. All the colours of the Mauritian rainbow are not present in their own rights. One is constantly reminded of the plurality and the diversity of Mauritius. Some political leaders spend their time telling us that there is no small community in our country. This description of plurality and diversity invariably hinges on more than four communities. Yet some of them are underrepresented politically as a result of their non inclusion as specific eligible groups for the purpose of the BLS. It should be pointed out that in some cases, the size of some of these excluded communities is much larger than that of a very small community, which is eligible separately for the BLS. If a system, however imperfect, is designed to correct for political underrepresentation on the basis of communities, then it should apply indiscriminately to all of them. The exclusion of some ethnic groups from the allocation of Best Losers has caused some communities to be politically underrepresented. On the basis of seven communities being eligible for the BLS and using some proxies to ascertain their respective numbers,the apportionment of the BLS would have been different. The composition and sequence of the communities (if there were seven or eight instead of four) receiving the Best Loser seats would have been very different. This makes a huge difference in terms of descriptive representation and the rainbowness of Parliament.
stakeholders, however well intentioned,
do not always have all the expert
knowledge and information to fully grasp all the
ramifications of different
electoral systems. They believe
what they know is what is best for the
country or what is best for them
is also best for the country.»
While some specific components of a large community have lost some degree of political and parliamentary representation because for the purpose of the appointment of Best Losers, they are considered as a whole,the case of creoles is dramatic. There has been a call from Père Grégoire and the FCM to amend the definition of the term ?General Population? and to divide it into two distinct components (Creole and others). The only reason for the presence of the four communities in the Constitution is the Best Loser. There is no other justification for such categorization. One wonders whether the intended consequence of the request of Père Grégoire and the FCM is to have a more realistic definition of the different components of the population of Mauritius so as to reverse the acute political underrepresentation of creole.The Best Loser System has been commented upon very unfouvarably by many Judges (Lallah, Glover, Seetulsing and Balancy). It must be pointed out that Judge Seetulsing reached the conclusion that the current definition of the four categories in the Constitution raises many problems for the BLS. While it is not difficult to ascertain (broadly as opposed to exactly) the level of underrepresentation of some communities ,the case of Creole is tricky unless we have a population census and ask the direct and open question of who is a creole. In many cases, it is obvious who is a creole in the GP configuration; equally in other cases, it is plain who is not a creole. However there are cases where only a specific declaration by the individual will determine the identity. Using some simple assumptions, I have tried to determine the potential political underrepresentation of creoles. Obviously the conclusions vary depending on the exact size of that specific community. However,even in the most conservative scenario, there is significant Parliamentary underrepresentation. One has to look at the current composition of our Parliament to realize it. If Creole were classified separately for the purpose of the Best Loser system, there would have been more creole in the National Assembly. There are, at least, five communities that are potentially, if not effectively underrepresented because of the discriminatory definition of eligible communities.
Often the allotment of BL seats leads to unequal distribution of MPs across constituencies. There have been cases of constituencies with 5 MPs instead of 3. In one case (1976) there were 6 MPs representing one particular constituency. It can also worsen the problem of malapportionment. Some constituencies are already overrepresented as they return three MPs with a small electorate. If in addition best losers are appointed from these same constituencies, their overrepresentation becomes greater. This happened in 2000 when best losers were awarded to members from two very small constituencies.
The number of electors in a constituency affects the allocation of BL seats. As a result, a candidate can secure a BL seat with far fewer votes than another one belonging to the same community and the same party. The most known example is the case of Rima and Koenig in 1967. Rima was alloted a BL seat with 6,596 votes while Koenig, the leader of the PMSD, did not have one with 6,851 votes. Both were from the PMSD and both belonged to the same community. It was the smaller size of the constituency that favoured Rima as his percentage of votes at 45.72 per cent was slightly higher than that of Koenig at 44.37 per cent. The case of Balancy and Chaperon in the same elections confirms how constituency size can influence the allocation of Best Loser seats. Both were from the Independence party and both were members of the General Population. Balancy became a Best Loser MP with 4,884 votes while Chaperon was not with 5,998 votes. The constituency of Chaperon was 34 per cent bigger than that of Balancy. In 1995 and 2000, it led to a double anomaly. In 1995, Beeharry, a Muslim, was awarded a BL seat even if his party won a very insignificant percentage of the national vote compared to the 20 per cent of the MSM/RMM. Worse he won only 4,405 votes while Soodhun, also a Muslim, polled 7,416 votes without a BL seat. In 2000, Leopold with only 7,732 votes (his party got 1 per cent of national vote) won a BL seat while Petit with 14,626 (and his party had 37 per cent of votes overall) did not.
It is possible to abuse the system. After an examination of the provisions of the Constitution, Judge Seetulsing reached the conclusion that it is perfectly constitutional for Seegobin, Lallah and Renghen (all Hindus) to be classified as General Population, which is a residual community in the Constitution. He has proposed that either the definition of eligible communities should be reviewed to ensure better classifications or the BLS should be done away with, because of its unenforceability. Consequently, he has questioned the enforceability of the BLS and stated that this problem should be addressed within the context of electoral and constitutional reform. In essence, there is nothing that prevents a candidate of a particular community choosing, for the purpose of elections, to become a member of another community so as to enhance his chances of obtaining a best loser seat. There are candidates that select their community for election purposes by drawing lots as they castigate the communalistic aspects of the Best Loser system. For instance candidates of the Chinese community, who are also Christians, could improve their chances of being a Best Loser MP if they chose to be classified as General Population. Such abuse could pervert the basis for which these seats are alloted. If, in the light of the Seetulsing judgement, S.Ghurburrun (a Hindu) had registered as a General Population in 2000 (or simply not declared himself as a Hindu), he would have been the fourth Best Loser from the second set of four with 43.99 % of the votes in lieu of XL Duval with 42.77 % .While being constitutional,this is simply absurd.
Archaic and anachronistic
In its current form, the BLS is archaic and anachronistic. Subsequent to a Constitutional amendment in 1982, there is no need for Mauritians to reveal their ethnic identities for the purpose of population censuses. Since 1972, there is thus no official publication of the ethnic make up of the population. And yet BL seats continue to be allocated according to the 1972 communal breakdown. For instance, the BL seats in 2005 were distributed using the population size and the communal breakdown of the 1972 census, almost 33 years before. The population has since increased by almost 40 per cent and it is highly likely that the ethnic composition is not the same as in 1972. The last population census prior to the one in 1972 was carried out in 1962 which obviously had a different ethnic mix. If within a period of ten years,the ethnic communal mix of the country had changed to affect the appointment of best loser seats,one can imagine what has happened over a period of 35 years. There was a higher percentage of General Population and Chinese in the 1962 census than in the 1972 one. I have run a detailed simulation of how the allocation of BL seats would have been affected if we were to use the 1962 as opposed to the 1972 survey. Simply amazing. Amongst others, P. Bérenger would have been a BL member in 1987 while S.Tang would have made it in 2000!
Yet the system continues to use outdated statistics to allot BL seats. The reliability of these figures and the fairness of the method can easily be contested. It could also lead to anomalies. For instance, if the 1962 census were used instead of the one of 1972, the allocation of BL seats would be different in terms of communities and individuals for most elections. Second if we use religion as a proxy for community in subsequent censuses and make some simple assumptions, this would give a third set of BL. Finally, if there were seven eligible communities instead of four, the allocation of BL would again be different.
Complex, arbitrary, erratic, unreliable, unpredictable
BL seats are allocated by the use of a very complicated formula using a two tiered system. Very few people understand how it operates as it has to take many factors into consideration. First the appropriate community, then the appropriate party with the appropriate community and finally it has to maintain a given political balance.
The process can be quite arbitrary in its apportionment. In 1995 a party that polled 20 per cent of the votes did not secure a single best loser seats. Four Best Loser seats were instead alloted to parties with a very small percentage of votes while the MSM/RMM had members of the appropriate community among its defeated candidates. In 2000 and 2005 also, two best loser seats were distributed to a very small regional party at the expense of a large national formation that had many candidates of underrepresented communities in its ranks.
The system is quite erratic in its distribution. In the 1983 General Elections, Nawoor with only 16 per cent of the vote was chosen as a Best Loser for the Muslim community while Kasenally with 47 per cent and belonging to the same community was not returned. Likewise in the 1987 elections when Finette with 47 per cent of the votes obtained a Best Loser seat while Bérenger did not go to Parliament with 49 per cent of the votes even if both are from the General Population. This is because of the specific complexities, randomness and the two tiered basis of the allocation method.
There is no by-election to replace a vacancy for a Best Loser seat. In case of a vacancy, the seat is alloted to the most successful unreturned candidate available, belonging to the same community and party as the outgoing member. This could lead to a situation where the candidate may have given up politics or have left the country. Worse, he could have changed sides in the meantime.
The system is also very unpredictable in its outcome. In 1982, 1991 and 1995 only four Best Loser Seats were distributed as the winning party did not have any unreturned eligible candidate. During those years, there was a Parliament of only 66 members instead of 70 MPs as provided for in the Constitution. As a matter of fact, the Electoral Supervisory Commission did not know whether to appoint any Best Loser in 1982 as it could upset the balance of political forces. The matter was referred to the Supreme Court that ruled that only the first four Best Losers be alloted.
The process can be mind boggling in some cases. In 1983 there was an alliance among three parties (MSM, LP and PMSD). However to avoid three symbols on the ballot paper, it was agreed that in most constituencies all candidates of that alliance would stand under the joint symbol of the sun and the key (MSM/LP) while in some districts the symbol would be a cockerel only (PMSD). Notwithstanding the fact that the three formations were in a total alliance yet they were registered as two parties for the purpose of the elections (as MSM/LP and as PMSD). For tactical reasons, Malherbe, who was a very long serving member of the LP, had to stand on the PMSD platform in the first constituency of Port Louis. He polled 44.91 per cent of the votes and yet was not returned as a Best loser while Candahoo, from the same alliance and belonging to the same community, got in with 43.78 per cent of the votes, as he contested under the MSM/LP banner! Had Malherbe stood under the symbol of his party, he would have been chosen as a Best Loser. In the same elections, Clifford, a PMSD candidate for the MSM/LP/PMSD alliance, polled 46.27 per cent of the votes; yet he was not awarded a Best Loser seat while Michel, from to the same community and the same alliance, got one such seat with 44.32 per cent of the votes as he fought the elections as a MSM/LP candidate.
The BLS allows communal parties to access the Assembly, as there is no threshold requirement. This is against the spirit of the framers of the Constitution who did not favour single issue parties and preferred broad church formations that straddle all communities. One such openly communal party did receive a Best Loser seat in 1995 at the expense of a national formation.
Banwell (1966) proposed the Best Loser system in a very specific context and it was intended as a purely temporary arrangement to facilitate the difficult political transition to the country?s independence. He stated optimistically that: ?we hope, however, that it will prove possible for all parties in Mauritius to agree on the ultimate disappearance of such political arrangements for communities?. (Banwell,1966,p.7) 42 years on the temporary arrangement is still operational.
Why does a system that is so criticised remain in place? Is it for lack of a better alternative? Even those who despise the communalistic aspects and the inequities of the Best Loser System acknowledge that it is imperative to have broad social and ethnic representation in the legislature to reflect the diversity of our country. Thus the Best Loser System becomes a necessary evil that plays an important inclusion role in society. Some argue for its maintenance until a better alternative is found. For them, the problem lies in finding a solution that both does away with the current formula and provides for broad ethnic representation. There are two issues at stake. First, the number of candidates from each community fielded by major parties and second, the likelihood of such candidates being elected. The major drawback of the current system is that parties can influence only the first issue as one cannot compel electors to vote for all three candidates of the same party. The Best Loser therefore acts as a second chance for some. There have been some instances of split votes along ethnic lines even if, in the overwhelming majority of cases, people vote for the three candidates of the same party. Short of a viable alternative that ensures broad representation, many stick with the devil they know. Under these circumstances, the bias is in favour of keeping the existing voting formula inspite of its shortcomings.
Sachs (2002) stated that there are four possible responses to the BLS. They are as follows:
● it can be abolished without more ado;
● it can be retained as is;
● it can be kept on in a modified form;
● it can be subsumed into the new dispensation .
Some of the political elites have decided not to abolish the BLS in spite of the fact that the great majority of deponents criticised it. Even those who defended the BLS stated that they did not like it and saw it as a necessary evil until a better alternative is found. The same elite has also not accepted the recommendations of the Sachs Commission to subsume the 8 Best Loser seats within the 30 PR seats notwithstanding that: ?the mixed system of 62 FPTP and 30 PR MPs introduces a degree of flexibility which can be advantageous to parties seeking to ensure that its leadership core will be returned to Parliament and it will also make it easier for parties to introduce predictable elements of community, religious and gender representation. This might reduce or eliminate the need to retain the BLS.? and ?there was widespread support for the idea of using the list system to provide for establishing mechanisms that will subsume the BLS and further its underlying affirmative objectives without perpetuating its anachronistic and divisive aspects.?
For experts in the political science of electoral systems ,this is so obvious as abundantly evidenced by what exists in South Africa and in other plural societies using PR party lists. Yet some are trying to be smarter than they ought to be. Some political stakeholders, however well intentioned, do not always have all the expert knowledge and information to fully grasp all the consequences and ramifications of different electoral systems. They believe what they know is what is best for the country or what is best for them is also best for the country. Under these circumstances the fate of reform is inevitably compromised. How sad from the very people who are desperate about reforming the electoral system. Hoist by their own petard!
By Rama SITHANEN