CHAPTER 7: THE SITHANEN REPORT

                                                                                                  white-paper-c7

Very soon after the publication of the Carcassonne Report, Dr Rama Sithanen published his latest report documenting proposed changes to the voting system in Mauritius. Dr Sithanen was the Minister of Finance of Mauritius between 1991 and 1995 and the Vice-Prime Minister and Minister of Finance and Economic Development between 2005 and 2010. Having taken an interest in reform of the electoral system in Mauritius for many years, Dr Sithanen was able to draw on his deep understanding and produce a detailed and comprehensive report in excess of 100 pages, which unlike any of the preceding reports tests the effect of its proposals using extensive polling data from previous elections.
The effects of different proposed electoral systems may not be fully understood by those involved, but reforms need to be understood with regard to the relevant contextual and temporal factors.67
67 R Sithanen, “Roadmap for a better balance between stability and fairness in the voting formula” (January 2012) p 23

Problems with the Carcassonne Report
Despite regarding the Carcassonne Report as having proposed a novel idea, the Sithanen Report concluded that it created too many practical problems and had too many unintended consequences. In particular, the Sithanen Report perceived the following to be weaknesses of Carcassonne's proposal:

(1) radical changes to the current First-Past-the-Post system that has been widely accepted in Mauritius for many years;

(2) absence of well defined boundaries for the multi-member constituencies, the redrawing of which would probably result in deep division and bitter disagreement;

(3) risk that no clear majority will emerge by reference to elections in Spain since the restoration of democracy in 1977, which would result in minority governments necessitating the support of many small parties to govern, despite Carcassonne's proposed measures to strengthen the system by, for example, proxy voting and making it more difficult to dislodge a sitting Prime Minister;
41
(4) danger of not meeting the criteria of diversity and inclusiveness at the constituency level, albeit that concern may be ameliorated at a national level;

(5) disenfranchisement of the elec-torate by allowing the political parties to decide who goes on the list and in what order; and

(6) weakening of the link between the elected representatives and their constituents as it currently exists in a First-Past-the-Post system.68

68 R Sithanen, “Roadmap for a better balance between stability and fairness in the voting formula” (January 2012) p xiii
69 R Sithanen, “Roadmap for a better balance between stability and fairness in the voting formula” (January 2012) p 3
70 R Sithanen, “Roadmap for a better balance between stability and fairness in the voting formula” (January 2012) pp 4-5
No need to throw away the baby with the bathwater
In respect of the Carcassonne Report, Dr Sithanen commented that there is, "absolutely no need to throw away the baby with the bath water".69 Mauritius had on various occasions considered the possibility of a fully proportional representation system, even as early as the First London Conference in 1956. However, Mauritius had decided upon the existing system as striking the correct balance between stability of government and fairness.70
Dr Sithanen created what he termed an "extremely bold attempt" at testing the Carcassonne model using some realistic assumptions about constituency boun-daries, geographical support of the two main parties and the likelihood of a third political formation fragmenting votes in few constituencies.71 The model he tested appeared to show that Carcassonne's proposals would not lead to a majority, despite the leading party capturing 49% of the vote. These concerns about a system that would not yield stable governments weighed heavily against the Carcassonne recommendations.
71 R Sithanen, “Roadmap for a better balance between stability and fairness in the voting formula” (January 2012) pp 15-16
First-Past-the-Post and stability
The main difference between the first-past-the-post system and the proportional representation system is that the former prioritises a stable government at the expense of fair representation, whilst the latter tends towards a fair representation of the votes in Parliament at the expense of stability. Dr Sithanen believed that the existing system needed reform to remedy the distortion between the percentage of votes achieved and the percentage of seats those votes translated into, without 42

unduly upsetting the stability of governments.72
72 R Sithanen, “Roadmap for a better balance between stability and fairness in the voting formula” (January 2012) p xiv
73 R Sithanen, “Roadmap for a better balance between stability and fairness in the voting formula” (January 2012) p xv

The proposal
The essence of the system proposed was, firstly, the retention of the three-member electoral constituencies, coupled with the maintenance of electoral boundaries and the acceptance of unequal distribution of voters across constituencies, and, secondly, an additional tier of 20 MPs combined with a closed, rank-based party list to return these MPs on the basis of otherwise wasted votes, subject to a threshold for eligibility, and the provision of double candidacies.73 Accordingly Parliament would have 82 MPs under the new proposal.
The under-representation of women in Parliament was a concern, which Dr Sithanen recommended be reformed by requiring parties to select no more than two candidates of the same sex in each of the twenty First-Past-the-Post constituencies in Mauritius, and if there are to be party lists, there should be one
person of a different gender out of every three sequential candidates.74
74 R Sithanen, “Roadmap for a better balance between stability and fairness in the voting formula” (January 2012) pp xiv and 34
75 R Sithanen, “Roadmap for a better balance between stability and fairness in the voting formula” (January 2012) pp 26-27
76 R Sithanen, “Roadmap for a better balance between stability and fairness in the voting formula” (January 2012) pp 27-28
Dr Sithanen grounded his proposals in what he saw as the six core values in a plural society such as Mauritius: (1) government stability; (2) party fairness; (3) broad-based socio-demographic inclusion; (4) gender representation; (5) accountability; and (6) avoidance of communal parties.75 Although it would be almost impossible for a single electoral formula to attain all these attributes, it was intended that the proposals would address the dis-proportionality between seats won and votes polled in Mauritius, not undermine the stability of the system, promote equal representation of the sexes, and find a viable alternative to the Best Loser System that ensures all sections of the population are adequately represented.76 43

Four special features ensuring repre-sentation of all segments of the population
Obviously a major factor that influenced the design of the proposals was the inclusion of different racial, ethnic, religious, linguistic and cultural groups in the electoral process. Dr Sithanen regarded the Mauritian electoral system as containing four special features that guaranteed the political representation of all the main segments of the population:
(1) specially drawn constituencies, following an intelligent pairing of Sir Malcolm Trustram Eve’s forty districts into twenty constituencies, to ensure broad representation;
(2) unequal distribution of popu-lations across electoral districts to encourage diversity and plurality, especially for groups dispersed across the country;
(3) three-member constituencies with a mandatory three votes per elector, recognised by the Sachs Commission as encouraging parties to straddle community divides when nominating candidates due to the political ad-vantages of a broadly based ticket; and
(4) the Best Loser System.
It was in that context of other safeguards being available to protect the interests of
all sections of society that Dr Sithanen felt able to recommend substantial reform of the Best Loser System.77
77 R Sithanen, “Roadmap for a better balance between stability and fairness in the voting formula” (January 2012) p xv
Dr Sithanen recognised that, while there is a historical and symbolic dimension to the Best Loser System, the specially drawn constituencies, unequal voters per constituency and three-member electoral districts all had a much more pronounced impact on the parliamentary repre-sentation of some components of the population than did the Best Loser System.
Problems of the Best Loser System
The Sithanen Report highlighted some of the main problems with the Best Loser System including that it is anachronistic, being based on the 1972 census; that it relies on a complex two-tiered system; that history is littered with what seem to be erratic results when a candidate with more votes than a colleague in the same community is rejected for not being in the “appropriate party”; there is scope for abuse where candidates claim that they belong to the General Population in order to position themselves for a best loser 44

seat78; and that it often does not result in all eight best loser seats being filled.79
78 See the judgment of Seetulsingh J in Carrimkhan v. Lew Chin and others (2000) MR 145
79 R Sithanen, “Roadmap for a better balance between stability and fairness in the voting formula” (January 2012) pp 51-55
80 R Sithanen, “Roadmap for a better balance between stability and fairness in the voting formula” (January 2012) p 60
81 R Sithanen, “Roadmap for a better balance between stability and fairness in the voting formula” (January 2012) p 86

Unreturned Votes Elect
Rather than using the parallel system referred to by the Select Committee which would not address the imbalance of votes and seats, or the compensatory mode of allocating proportional representation seats favoured by the Sachs Commission, which converts the system into a proportional one in many circumstances with the attendant risks of instability, Sithanen proposed the “unreturned votes elect” formula to elect the 20 list tier MPs, i.e. the use of votes of candidates who had polled well but had not been elected.80 The votes of unelected candidates of all parties are taken into account in all 21 constituencies. The votes are aggregated and the parties are allocated seats according to their share of votes of unreturned candidates nationally.81


Simulations
Various simulations were conducted by Dr Sithanen to test how the various systems might have operated in past elections. A good summary of some of those analyses is provided where Dr Sithanen writes,
As expected, the FPTP produces the largest majority in all elections. It ranges from 14 seats in 2005 to 54 seats in 1982 and 1995. Sachs Model C lowers considerably the majority in 5 elections; it varies from no majority in 1987, to a majority of 2 seats, 4 seats, 4 seats and 8 seats in 1987, 1983, 2005, 2010 and 1967 respectively. The UVE formula sits between FPTP and Sachs Model C (closer to the latter) in the five hotly contested elections of 1967, 1983, 1987, 2005 and 2010 with a majority varying from 10 seats in 1967 and 2005, 12 seats in 1983 and 2010 and 8 seats in 1987. Essentially it gives a working margin to the winner. In 1995, it delivers all 20 seats to the MSM while Sachs operates slightly differently. The 1976 elections delivered a hung Parliament in all three modes.82
82 R Sithanen, “Roadmap for a better balance between stability and fairness in the voting formula” (January 2012) p 95