Thoughts on Electoral Reform (4)

At 60 miles an hour the loudest noise in this new Rolls-Royce comes from the electric clock.

David Ogilvy headline

We Already Have the Rolls-Royce of Constitutions

Our constitution through its excellent First Past The Post (FPTP) system has enabled voters to proceed with five seamless changes of government since independence. Three of which happened after consecutive elections. And most of our governments have been stable. This has been a key ingredient for our early successes. Our FPTP system has allowed us to focus on solving difficult problems which many back in the 1960s had rightly considered as almost unsurmountable. We also appreciate the quality of our constitution when political alliances break down: new ministers are appointed swiftly. And when we see how the absence of a good enough constitution in other countries hold them back. But it's not perfect. It has a slight clicking sound. Which needs to be corrected soon. In a mindful fashion.

But it Has an Electric Clock

Essentially our electoral setup rewards winners a little too well. And there have been a couple of times when the results have been extreme. Like in 1982 and 1995 when one alliance swept all the seats. These two electoral outcomes have highlighted a problem with our FPTP system: we could have an opposition that's a bit too small. Indeed as Table 1 illustrates we ended up with an opposition of only four best losers in 1982 after one alliance won all the regular seats. In fact this has more to do with the way the current best loser algorithm allocates additional MPs than with the FPTP system per se. So how do we fix this? For sure we know what we don't want.

No to PR and Party Lists

For at least three reasons. One is that the party winning a majority of seats may end up not forming government. Like we've seen happening recently in Rodrigues. Two is that MPs will stop worrying about voters and instead focus where they are likely to land on party lists. This will mean bending over backwards to please party leaders in any which way. And three party lists will take away one of our most important prerogatives namely that of deciding which candidates to keep out of Parliament. This is an important tool that we have to refresh the slowly-changing parts of our political scene. We shouldn't also forget that our Parliament is already way too big: a couple of years ago Mauritius ranked 149th in terms of population/MP (thank you Wikipedia). So that suggestions to increase its size by as much as one-third is kind of ridiculous.

Quieting the Electric Clock

If we're discarding the tortuous best loser allocation procedure and we don't want party leaders to step in for obvious reasons then how do we solve this? By doing two things. One is to fix the minimum size of our opposition and the other is to devise a rule to pick the extra MPs. First let's tackle the minimum opposition size.

The parameters which need be considered here are (i) again, that our Parliament is already too big; (ii) we should move back to 40-odd single-member constituencies (SMCs) like we had in 1959 and 1963 as they offer plenty of advantages; (iii) the maximum of best losers currently allowed (8); and (iv) the share of votes obtained by the party/alliance which had the second largest total number of votes in the two clean sweeps of 1982 and 1995 (these were about 1/4 and 1/5 respectively).

It turns out that it makes a lot of sense to keep the maximum of best losers at eight. Especially when we factor in what we know about the optimal size of groups that put dissent at a premium. Furthermore if we get a 42-0 in an SMC setup eight best losers would represent 16% of the total seats which is quite close to the 1/5 of votes obtained by the losing alliance in 1995. Part of the focus of the approach proposed here is to always have a functional opposition of a minimum of eight MPs.

How Would We Pick the Extra MPs Then?

Quite simply in fact. See, all that we would have to do is to select up to eight of the candidates that came right after the last person to be elected -- we would use the percentage of votes received as voter population across ridings varies quite a bit. Table 2 shows what our best losers would have looked like if we applied our simple and transparent rule.

Ramgoolam would have gone in along with Jagatsingh, Boolell and a few others. Duval would go in here too but Nicol Francois wouldn't if a little tweak to account for the relatively much larger voter base in Mauritius was applied: if both elected MPs of Rodrigues are from the same party our sister island gets only one best loser and if they are from different parties she gets none.

Comparing Tables 1 and 2 clearly illustrates how voters do a different job of picking best losers. Indeed the simple rule would have produced an opposition that's two times bigger and represents almost three times more votes. This of course would have reduced the disproportionality of the FPTP system in that particular year without jeopardising the stability of the outcome. The two tables also indicate what we lost in terms of quality of opposition. Because of an algorithm which is based on ethnicity.

What About the Second 60-0?

Table 3 shows the four best losers we got while Table 4 contains the results after using the simple rule proposed. Jugnauth, Bachoo and Sithanen would have made it.

I don't think I am making a big mistake in assuming that many who supported the MMM/PSM alliance in 1982 would disagree that our Parliament would have worked a lot better having the eight best losers in Table 2. Just like right-thinking supporters of the LP/MMM alliance will agree that having Jugnauth and Sithanen join Duval in the opposition in 1995 would have been a lot better than what we got. Interestingly while outgoing Prime Ministers would have been the first to get best loser seats after the two 60-0s our simple rule would not have allocated any additional seats to the island of Mauritius in 2014 -- and therefore would not have saved Ramgoolam, Boolell or Sithanen -- because we already had an opposition bigger than 8 MPs. Similarly Berenger would have stayed out of Parliament in 1983 as voters returned an opposition that was even larger.

One thing though. Votes of this variety of best losers (aka opposition MPs) should not count. Because that could introduce way too much instability especially when a two-party alliance who had split electoral tickets in two equal halves wins a supermajority and then breaks down. But they would be allowed to vote in Parliament because that would help us find out where they stand on important issues. Likewise they shouldn't be eligible to join Cabinet.

Simple Rule Solves Another Problem Too

That of satisfying the UNHRC ruling as it doesn't require to stamp us voters as cattle. So to recap, we proposed a simple rule which will allocate up to eight best losers in an efficient manner. It doesn't make use of party lists which are inherently dangerous as they suppress dissent. It will not make our relatively huge Parliament any bigger. In fact it has been calibrated to work even better when the Electoral Boundaries Commission eventually slices Mauritius back into 40-odd single-member constituencies. It's a rule that everybody can understand. And uses the expressed preferences of increasingly smart voters to pick additional MPs when that's needed. It poses absolutely no risk to outcomes such as those which happened recently in Rodrigues where a formula frustrated the overall wish of voters. But it will make sure that the additional MPs help us have an opposition of sufficient size at all times. In other words it will quiet the electric clock of our constitution. And this will deepen our democracy.

Read part 1, part 2, part 3, part 5, part 6, part 7, part 8 and part 9 of this series.