ANALYSIS: PR shortcomings in Rodrigues will not occur in Mauritius

The 4-seats advantage at the constituency or FPTP level, which vanished upon the application of Proportional Representation (PR) in the just-completed elections in Rodrigues, is due to the confluence of 3 factors. In this article, I will highlight them, advise how they can be mitigated and finally assure all that they are unlikely to occur in Mauritius.  
Any implementation of PR in Rodrigues, whether it is the current format similar to Sachs Model C or the innovative Unreturned Votes Elect concept proposed by Rama Sithanen, faces 3 challenges.  
First, there are too few seats in contention. With only 18 seats, each seat represents about 100% / 18 or 5.5 % of the votes. If the vote split is say 53% v/s 47%, the seat advantage may just be 1 or none. By contrast with a potential of 82 seats overall in Mauritius, a 53% v/s 47% split will translate into a comfortable majority of 5 seats or more.  
Second, the election results have historically been very close e.g. about 55% v/s 45% in the 2002 and 2006 elections which translates into a 2-seat advantage in a legislature of 18 members once PR is applied. The difference in votes between OPR and MR, either at the constituency or the Party List or Regional level, was even closer in the election which just concluded.  Consequently, the 8-4 FPTP result flattered OPR. For instance, OPR and MR could have easily split the 2 seats in St Gabriel, thus bringing the FPTP score to a more equitable 7-5 or just a 2-seats advantage!
Third, the usage of 2 sets of votes – one at the constituency level and another at the Party List or Regional level can lead to result “inconsistencies”. Voting for Party A candidates at the constituency level and simultaneously for Party B at the Party or Regional level is perfectly legitimate but it can create an “inconsistency” in the translation of votes into PR seats.  
If only the votes cast for constituency candidates were considered in the PR exercise, FPR would have tallied only 8.4% of the votes. Thus, it would NOT have met the 10% threshold for PR and thereby, it would not have been allocated any PR seat. The overall seat allocation would then have been 10 for OPR and 8 for MR.
With 2 sets of votes, many voters gave a “consolation” vote to FPR and in the process, it collected about 10.6% of the votes at the Party List / Regional level. Thus, FPR got 2 PR seats as 10.6% of 18 rounds up to 2. The compensatory PR exercise further allocated the remaining 4 PR seats to MR. With the seat split then at 8 each for OPR and MR and 2 for FPR, the Electoral Commission was forced to allocate 3 more PR seats to OPR to re-establish its FPTP dominance.
It was indeed a confluence of these 3 shortcomings which led to this rather controversial outcome. Rodrigues is perhaps better off with ALL parties working together (i.e. unity council) in view of the close votes.
Two quick fixes to make PR more effective in Rodrigues would be
1.    STOP requiring voters to cast a second vote for a Party to avoid potential inconsistencies, such as the one that just arose with FPR.
2.    CUT down the PR seats from 6 to 3 (and thereby, the Regional Assembly caucus from 18 to 15) in order to make the seat outcome less proportional to the votes. The mathematical logic is that 6 PR seats out of 18 (or 33%) brings in too much proportionality in the final seat allocation. In Mauritius, we are talking of only 20 PR seats out of 82 (or 24%). Additionally, an odd number of overall seats (e.g. 15 in Rodrigues) is more conducive to a majority outcome. For the 2012 election, the final seat allocation would have then been OPR 8 MR 6 FPR 1. OPR would have had an outright majority without the need for additional PR seats.
The above PR shortcomings in Rodrigues are unlikely to happen in Mauritius because with  82 seats potentially compared to 18 in Rodrigues, the difference in seat allocation will be magnified 4.5 (=82 / 18) times more in Mauritius compared to Rodrigues. It will be easier to enjoy a bigger seat advantage in Mauritius despite close vote percentages.
Switching to a PR electoral system requires a new mind-set among all stakeholders. Specifically, a party can be “guaranteed” a win only if its constituency or FPTP seats tally to more than half of overall number of seats (i.e. 10 out of 18 in Rodrigues and 42 out of potentially 82 in Mauritius). Otherwise, since PR seat allocation is calculated on the percentage of votes, it is this statistic which matters more than the FPTP seats secured. Here's an illustration for Rodrigues. Suppose Party A barely wins 8 constituency (or FPTP) seats. Party B is highly competitive across all 6 constituencies in Rodrigues but it only wins 4 constituency seats. It is very plausible Party A secures only 43% of the (second) Party List / Regional votes compared to 57% for Party B. When the PR correction takes place according to the D'Hondt Highest Average Method (an online calculator is available at http://icon.cat/util/elections) employed by the Electoral Commission, Party B will be awarded all the 6 PR seats and win end up the overall winner with 10 seats.  
Think about PR as a soccer game with 2 halves. The recent Chelsea – Man United 3-3 game comes to mind. The team (party) leading at half time (FPTP seats) will not necessarily win the game (election)!