“We have to make choices. Singapore chose to live with little or no opposition. Are we willing to accept that in Mauritius?

Interview: Prof Singfat Chu, Professor of Business Analytics -National University of Singapore

‘many are voicing loudly for “equal” constituency sizes unaware of the clear danger that in trying to be equal, fewer minority candidates would be elected through FPTP and BLS. Our representation may end up more unequal’


Prof Singfat Chu, of Mauritian origin, has been a Professor of Business Analytics at the Business School, National University of Singapore since 1991. He has penned many articles in the local media on the analytics of electoral reform in Mauritius. Here he gives us his contribution to the ongoing debate about electoral reform. He proposes a model that, according to him, would achieve more equitable representation without causing much disruption and without increasing the number of MPs. Read on:


* Mauritius Times: What’s your take on the electoral reform proposals announced by the Government last Friday? Prof Singfat Chu: First of all, let me say that according to (Administration and Cost of Elections) Electoral Knowledge Network,[1] “electoral reform is a broad term that covers, among other things, improving the responsiveness of electoral processes to public desires and expectations”. Two fundamental “principles” outlined by the government are simply not conducive to “…improving the responsiveness of electoral processes to public desires and expectations.” The fundamental “principle” of maintaining the First Past the Post (FPTP) seat advantage goes against the ethos of electoral reform. FPTP is a man-made device which has distorted election results at the expense of losing parties in the zero-sum game of seat allocation. The intent not to alter the FPTP seat advantage implies logically and responsibly that no extra seats should even be considered in order to save on public resources. The other mentioned principle of “government stability” should cogently be based on the “voeu de l’électorat” which is reflected in the percentage of votes obtained and not on distorted FPTP seat allocation. In fact, the shortcomings in our current electoral system can be traced back to the three (3) key solutions implemented in 1967. However pragmatic they were to allay all types of fears then and to enable us to survive the post-Independence journey, they have however led to the sprouting of many issues in the past 50 years. * Are you saying that the problems that are said to have emerged today can be traced to the solutions which the framers of our Constitution – and of our electoral system – proposed 50 years ago , but which have served us well down the years? Yes, indeed. Let me explain. The first of these 3 key solutions can be thought of as the division of a family home into 20 rooms (i.e. constituencies) in Mauritius and an “outbuilding” in Rodrigues. I will restrict myself to the 20 rooms in Mauritius. Each of them has elected 3 representatives to the “National Assembly” in each of the 11 elections held between 1967 to 2014. In those years, the number of voters has grown from 266,518 to 666,936 or by 2.5 times. Growth aside, a bigger problem has been the increasing disparity in the number of voters across the 20 constituencies whose geographical delimitations have hardly changed. The ratio of the maximum to minimum number of voters across constituencies which was 1.67 in 1967 has grown to 2.85 in 2014, a situation which is attracting louder grumbles. More comprehensively, the variability of voters across the 20 constituencies measured by the Gini coefficient has doubled from 0.06 (with 0 indicating the same number of voters by constituency) in 1967 to almost 0.13 in 2014. The different expansions of the constituencies have compromised a tenet of democracy namely “one person one vote” as the weight of a vote towards a seat differs according to constituency size. * Should we therefore go for a redefinition of our electoral boundaries, as it is being canvassed by the PMSD? Equalizing the number of voters across the constituencies is a quasi-impossible task. In fact, the geographical boundaries of constituencies have hardly changed since 1967. There is a delicate reason to that. Some small constituencies have deliberately been drawn as such to favour the election of minority candidates. Without this mechanism, minority candidates would hardly figure among FPTP elects and the Best Loser System (BLS) would not have redressed inadequacies. In fact, many are voicing loudly for “equal” constituency sizes unaware of the clear danger that in trying to be equal, fewer minority candidates would be elected through FPTP and BLS. Our representation may end up more unequal. * What about the second shortcoming, as you say, in our electoral system? The second solution of the framers of our electoral system was the choice of the majority or “First Past the Post” (FPTP) vote system. While this has been conducive to governmental stability, some would even say over-stability, it has been too much at the expense of the Opposition in the zero-sum game of FPTP seat allocation. In the 11 elections held between 1967 and 2014, the winning parties (pre- or post-election) have averaged about 55% of the votes (min=49.4%, max=66.2%) and about 46 or 77% of the FPTP seats (min=55% for post coalition government in 1976 and max=100% in 1982 and 1995). This average 22% seats-to-votes discrepancy compromises another tenet of democracy i.e. equality of representation. An over-represented majority implies an under-represented opposition. On top of that, another facet of non-representation in a FPTP system arises from the un-returned or “wasted” votes cumulated by the unsuccessful candidates. In the recent 2014 election, unsuccessful candidates cumulated about 46% of the votes cast. In 11 elections since 1967, about 43% of votes were un-returned or “wasted”. Even after the allocation of BLS seats, about 40% of votes cast had no representation in our National Assembly. Is this democracy? * This begs the question: has our majority based electoral system done a disservice to the country? I mean the majority rule system, tempered by the Best Loser System to provide for adequate minority representation? The answer to your question relates to the third solution namely the Best Loser System (BLS) which assuaged the family members (in the 20-room home) in 1967. I concur that in our initial decades of Independence, it has stood as a vanguard of adequate representation of the 4 communities named in our Constitution. The impact though has been more psychological than statistical. In the 10 elections from 1967 to 2010, the split of the 620 FPTP seats among the 4 communities was 59.4%, 25.2%, 13.7% and 1.8%. After the allocation of 67 BLS seats, the split became 54.1%, 29.4%, 14.8% and 1.6%. Grosso modo, the statistical impact of BLS has been light. Worse, BLS is based on obsolete 1972 census data. What I am saying is that, in summary, three solutions as proposed by the framers of our Constitution, which were pragmatic in 1967, have evolved negatively into problems. Now, as David R. Hawkins, author of ‘Healing and Recovery’, would say: “Problems are best solved not on the level where they appear to occur but on the next level above them… Problems are best solved by transcending them and looking at them from a higher viewpoint. At the higher level, the problems automatically resolve themselves because of that shift in point of view, or one might see there was no problem at all.” * What are you therefore proposing in terms of alternative and less disruptive reforms to our electoral system? Albert Einstein said, “We can’t solve problems with the same level of thinking that created them.” I am therefore proposing a simple yet effective disruption to our electoral system which will address fair translation of votes into seats and, collaterally, guaranteed stability, broad community, gender and other representations without additional seats. A momentous event in the life of a family is when it may have to decide whether to repair the family home with increasing cracks or to re-configure it. Inspired by the quote of Richard Hawkins, I believe that re-configuration is the pragmatic approach as we enter the second half century of our existence which I think should emphasize togetherness i.e. nationhood. The simplest and most basic re-configuration may well address many issues that we face currently. The proposal is to tear down the walls and resolve issues pertaining to constituency malapportionment and collaterally, FPTP and BLS through a switch to a Closed Party List electoral system with 2 doses of disproportionality to (1) deter undesirable parties, and (2) provide a “working majority” to a party with say at least 40% of the votes. The number of seats allocated will be at least 60 and at most 67 when providing for a “working majority”. We can decide whether the Party Lists must have all candidates ranked or have a mix of ranked and un-ranked candidates. An analogy is a soccer team with a starting line-up (i.e. ranked candidates) and thereafter the coach (i.e. party leader), can select among the substitutes (i.e. un-ranked candidates). The flexibility of a mix of ranked and unranked candidates is appealing. Casting a single vote for a Party represents a minor disruption to our electoral system as a cursory inspection of constituency results in all elections to date reveals clusters of 3 votes to the same party i.e. votes according to party affinity rather than for candidates. A Party List electoral system offers many attractions, namely:

  1. Every vote counts except those cast for parties which do not reach the threshold level for seats. Cutting down the current 40% wasted votes dramatically towards 0% is enlightened democracy.
  2. Fewer or up to the same number of seats as in the current system. This will address concerns about the higher number of elects e.g. via a mixed FPTP-PR system with say 80+ seats.
  3. Every segment of society must feature transparently in a competitive Party List. There are infinite combinations to achieve representative lists. The simple strategy of placing more minority, female, etc., candidates at the top of Party Lists will automatically take care of many apprehensions. Specifically, BLS will be subsumed and even improved via “common sense” rather than via a rigid framework.
  4. The elimination of “transfuges” as a seat belongs to the party rather than the holder. Any elect who leaves a party must also leave the National Assembly and will be replaced by another from the Party List. This takes care of “stability” and “governability”. A government does not need a big buffer of seats anymore. Simulations reported below indicate an average “working majority” of 11 seats in past elections.
  5. The emergence of more political figures at the national level as this is the platform where elections will be fought. Going forward, Mauritius needs more of such national level opinion leaders and doers.
  6. With up to 67 Party List seats in Mauritius and another 3 in Rodrigues, amending our Constitution will require 53 votes (=¾ x 70) in the National Assembly. A party is most unlikely to do it alone as it can only have these 53 seats with 89% of the votes. Support from other parties would be needed and that alignment is most welcome when making such fundamental amendments. This is democracy!

* Do you mean to say that the Party List, whereby party leaders obtain a second lever of influence — and decision — over who gets to contest elections would amount to democracy? Criticisms of the Party List system have centred on the weak links between voters and elects and the power of party establishment in selecting and ranking the candidates. I think both will be self-resolved as (1) post election, the elects will be divided to look after specific geographic areas, and (2) if candidates are not properly selected and ranked, votes may simply not be cast for the Party List and the party establishment will have to answer for that. The pressure is on parties to come up with lists which meet the expectations of the voters. Party lists are used solely or in a mix with other electoral systems in over 80 countries. The few cases which critics highlight ad nauseam are those which are too proportional , or have either no or a low threshold of votes for seat allocation. My proposal mitigates these. There is general agreement for a threshold in Mauritius. On top on that, I am also facilitating a “working majority”. I have simulated the Party List system with 2 doses of disproportionality namely:

  • Threshold of 5% (no issue with higher level) of votes to obtain seats;
  • If after the initial allocation of 60 seats, the “working majority” is 9 (this can be altered) seats or less, parties with 40% (i.e. a “contender to form the government”; this threshold can be altered too) will be considered for “bonus” seats. If there is only one such party, it will be allocated seats up to a maximum “working majority” of 10. If 2 parties meet the criteria, they will share 7 seats proportionately according to their votes in excess of 40%.

The simulation outcomes of this proposal in the past 11 elections appear in the Table I am submitting for your readers’ attention. Beyond the top 2 parties, the 5% threshold was breached only 3 times namely by PMSD in 1976 and 1982 and Parti Gaetan Duval in 1995. Bonus seats would not have been awarded in the three-way election of 1976 as no party breached 40% of votes as well as in the lopsided victories in 1982, 1991, 1995 and 2000 when a “working majority” of 10 seats or more emerged from the 60 initial seats allocated. In 1987, MSM-PTr-PMSD secured 49.9% of the votes compared to 48.1% for MMM-MTD-FTS. The initial 60 seats would have been split 31 to 29. As both parties scored above 40% of votes, MSM-PTr-PMSD would have been allocated 9.9 / (9.9 + 8.1) of the 7 bonus seats i.e. 4 more seats compared to 3 for MMM-MTD-FTS. This would have increased its “working majority” to 3 seats. This would have been just given the tight vote situation. In the recent 2014 election, L’Alliance Lepep with 50.8% of the votes would have been initially awarded 34 seats. PTr-MMM with 39.3% of the votes would have obtained the remaining 26 seats. Riding on its votes above 40%, L’Alliance Lepep would have seen its 8 seats advantage upgraded by 2 “bonus” seats to bring its ”working majority” to a more comfortable 10 seats. The “working majority” under this proposal would have ranged between 3 (in 1987) and 26 (in 1995) with an average of about 11 seats in the past elections. With “transfuges” no more a threat to “stability”, this would have been a comfortable buffer for the government. Even if a minor partner in the government were to quit (we have seen these following the 1982, 1995, 2010 and 2014 elections), its elects would have to vacate their seats and they would be replaced by those who were not returned from the Party List. New elections would hardly be needed and the government stability would therefore not be compromised. With autonomy in Rodrigues, it is not fair to include it in the proposed Party List system in Mauritius. A FPTP system remains the most practical system to elect its proposed 3 representatives to the National Assembly. A Party List system would not be incisive for seat allocation in light of the very small number of representatives. For example, each of the 60 party list seats in Mauritius would represent about 1.6% of votes compared to 33.3% in Rodrigues. There is no problem with a Party List system in Mauritius co-existing with FPTP in Rodrigues. * How do your proposals differ from the Singapore model, which like that of most Commonwealth nations, is modelled on the British Westminster system of parliamentary democracy and which is mostly FPTP based? Singapore currently has 13 Single Member Constituencies (SMC) and 16 Group Representation Constituencies (GRC) with between 4 to 6 seats. The concept of GRC, which is unique to Singapore and which was introduced in 1988, is to guarantee adequate representation of the Malay, Indian and Eurasian minorities in its Parliament. The contest in a GRC is among parties which must each field a team with at least one “specified” minority candidate. Voters cast a single vote for a party. For example, the Aljunied GRC is a 5 seats constituency which is designated to have at least 1 Malay MP. The electoral outcome in a GRC is obviously a whitewash (e.g. 5-0 in Aljunied) as the winning party gets all the seats. Since 1988, the dominant party has won in all the GRCs except on two occasions. The votes to seats discrepancy in Singapore is actually worse than in Mauritius. But with the GRC, there is no need for something like the BLS to have adequate minority representation. The price to pay for this is little to no Opposition representation. Currently, the Parliament in Singapore has 89 seats with 83 belonging to the dominant party. The Singapore Constitution states that there must be at least 9 Opposition members. “Best vote percentage losers” are selected to meet this requirement and they are labelled Non Constituency MP (NCMP). Another group of 9 known as Nominated MP (NMP) are selected from names submitted by civic groups such as academia, arts, labour movement, community service, etc. NCMPs and NMPs can partake in all debates but they cannot vote on sensitive bills e.g. the Budget, amendments to Constitution, recall of the President, etc. * Here in Mauritius, if we want to accommodate the concerns of both those who would like to see the BLS remain untouched as well as those who would who argue in favour of the decommunalisation of politics, we could just as well go permanently for an amendment of the Constitution which would remove the requirement of community declaration and ensure that the Electoral Commission nominates “Best Losers”. But can’t party leaders be relied upon for the fair representation of minority MPs? Removing the requirement of community declaration and yet maintaining some form of BLS based on “community” is a “deux poids deux mesures” oxymoron, isn’t it? I warned in 2014 that this “temporary” measure could turn out to be permanent. In using the average historical representation of communities to allocate seats, BLS is now perpetuating its own shortcomings. It takes two to tango for the fair representation of minority MPs. My analyses of past elections is that parties do their best in proposing candidates. Statistics stated above on the light impact of BLS supports this. The light redress BLS has made resulted from the “couper trancher” of some voters. It is too easy to blame parties. Voters must assume their responsibility too. * To come back to Singapore, how does its electoral system and it political leadership ensure that the country gets the political representation that has empowered it to move forward and achieve so much progress as compared to Mauritius? Recently, Pere Maurice Labour wrote in a newspaper to highlight the harmonious social scene in Singapore where for example communities live together according to set quotas in any public housing estate. I responded by highlighting that with the GRC concept, politics here are not divisive with respect to communities. In return, the government can implement such social projects. But there is a cost to that, namely as stated earlier, rare to nil opposition. This is why Singapore is often branded a “dictatorship” by the uninformed. We have to make choices. Singapore chose to live with little or no opposition and its powerful government has made things happen. Are we willing to accept that in Mauritius? There is a price to pay for every success. Mauritius has to decide which price it is willing to pay and for what. * Here, some parties are campaigning in favour of “Recall of MPs” legislation. This may not be available in the Singapore model, but how does the political leadership ensure that MPs do not lose their bearings? In my 26 years living in Singapore, 2 MPs including one who was the Speaker had to step down “on the spot” when they were found to have extra-marital affairs. Elections were held immediately to replace them. The culture here is that by virtue of their position, office bearers must be role models and subject to the highest standards of morality and integrity. It is too easy to clamour for “Recall of MPs”. Voters should look at the mirror and ask themselves why they voted for those MPs to begin with. Did they vote responsibly or for vested interests and found themselves with a “lemon”? I listen to radios and read newspapers everyday and sadly too many are quick to blame others rather than themselves. It is high time for voters to assume their responsibilities. If they did, there would be no need to recall MPs.


 * Published in print edition on 28 September 2018

Add a Comment

Your email address will not be published.