Some years ago, in a paper presented to the British Institute of International and Comparative Law, I argued that our electoral model, with its Best loser corrective, reflected a careful and delicate attempt at balancing several competing imperatives to allow the emergence of a stable and unified state from a diverse societal base.

I concluded that any attempt to substitute this model would have to begin with a proper appreciation of how difficult a tight rope balancing exercise it was. Reform, if it is to be undertaken at all, requires that we identify with precision the defects we claim to be correcting before we rush to redesign the system.

There is indeed a fundamental rule that we often invoke when considering legislative change, including constitutional change. It is called the Mischief Rule. The rule requires, first, that we identify the mischief, that is, the defect requiring remedy. It then requires that the proposed remedy does not create more problems than it seeks to resolve.

Now, there are two supposed defects or mischiefs that are often brought up to justify the claims for electoral reform. The first is that the electoral model can yield disproportionate results, with seats in the National Assembly not reflecting the proportion of votes obtained. The second is the consecration of communal considerations through the best loser system.

The point of this article is to invite readers to pause and reflect on these claims. Before accepting that reform is due, and before embracing solutions such as proportional representation as selfevident answers, we must ask whether we have correctly identified the mischiefs, and whether their remedy necessarily lies in electoral reform. n

Disproportionate representation

Following the last general elections, Alliance du Changement obtained approximately 96% of the seats in the National Assembly with about 62% of the votes, whilst the Alliance Lepep, with its 27% of the votes, only obtained about 3% of the seats, and this too, after the best loser corrective was applied. Previous elections have shown similar disparities. In the 2000 general elections, for instance, the MSM/ MMM coalition secured 88% of the seats with approximately 51% of the votes, while the MLP won 12% of the seats despite having about 36% of the votes.

Plainly, there is a disproportionate yield of seats to votes, but this is a known and well understood feature of the firstpast-the-post electoral model, and it is not, in itself, inherently wrong.

The issue or question is whether such disparity matters in any constitutionally meaningful sense. And if it does, whether the remedy necessarily lies in electoral reform.

One way to think about this is to ask whether it would really matter, for instance, that Alliance Lepep is not better represented in the National Assembly presently?

I have not seen this claim made with any real force. That is not to deny that a stronger opposition presence can always sharpen the functioning of the Assembly. The reality is that ordinary legislation is passed by simple majority in any event. From that perspective, it would make little difference if, say, 6 or 7 additional seats (where 20 additional seats were to be made available on a proportional representation basis) were allocated to the opposition in the National Assembly.

Nor can the issue be framed as one of constituency representation. Every constituency returns three members, and every constituency is represented in Parliament. Elected members represent all their constituents without distinction.

When calls for proportionality are made, the underlying concern is that electoral outcomes may allow a government to acquire the power to amend the Constitution without having secured a proportion of votes that may be accepted as providing such a mandate. The mischief, therefore, is not disproportionality per se. Rather, the mischief lies in the risk that constitutional amendment powers may be exercised on the basis of a parliamentary majority that does not mirror the popular vote.

Is that, however, a purely electoral issue? It is not.

Our constitutional framework, for instance, places limits on legislative power. Section 1 of the Constitution, which entrenches democracy itself, cannot be amended without approval by a referendum and unanimity in Parliament. This is a form of ex ante control over the powers of constitutional amendments. There is also ex-post control over such a power. An amendment to the Constitution may itself be declared unconstitutional by the Supreme Court if it is found to be in breach of Section 1. In 2016, when the then Government sought to introduce the abhorrent Prosecution Commission, we were prepared to seek redress from the Supreme Court for the Director of Public Prosecutions (DPP). Ultimately, the PMSD, in a commendable show of a constitutional conscience, rejected the Prosecution Commission and left power.

One may of course legitimately ask whether such safeguards are adequate. Even if they are not, the remedy may be simply to strengthen these ex ante and ex post controls. Because the reality is that there is always the possibility, under whatever model, that a party may secure a constitutional majority, and abuse it. A change in an electoral system does not change this, and the answer cannot be that no party in power should ever have a constitutional majority.

Over the last decade, we have fought off attacks on the Director of Public Prosecutions (DPP) and his office through constitutional litigation. Likewise, in the Valayden v The State, which is now before the Privy Council, we sought to challenge Parliament’s postponement of municipal elections on the ground that Section 1 imposes substantive limits on Parliament’s ordinary legislative power when democracy (through the holding of Municipal elections) is at stake.

Whilst, constitutional litigation presently takes time, there is no reason for it to be so. This mischief could be remedied by the setting up of a Constitutional court providing a streamlined process for urgent adjudication of constitutional questions. Such a safeguard may offer far greater protection than any adjustment to seat allocation.

What all of this points to is that the remedy of the supposed mischief may not lie in electoral reform, but elsewhere. Rather, it lies in strengthening the democratic and constitutional culture.

It lies in reinforcing constitutional safeguards, such as the independence of the Judiciary and the DPP, fundamental rights, access to courts, including the Privy Council, and accelerating the speed of constitutional adjudication.

A good constitution strikes the right balance. At present, the balance is between the disproportionate allocation of power and the ex ante and ex post controls that guard against its misuse. Our lived constitutional experience over nearly sixty years, which has been marked by stability rather than constitutional derailments, suggests that we have, by and large, got this balance right. It thus calls more for careful fine-tuning through the strengthening of safeguards than for a fundamental rethink.

What this also tells us is that an electoral system or model, and indeed its reform, cannot be considered in isolation. A constitution must be looked at holistically, including by looking at the intangible elements, such as the constitutional culture that has developed and been entrenched over the years.

The question, therefore, is not simply whether disproportionality exists. It plainly does. The question is whether disproportionality constitutes a constitutional mischief of such magnitude that its remedy necessarily lies in electoral reform. In view of the above, it does not.

Communal representation

The second alleged mischief is said to arise from the consecration of communal considerations through the ‘best loser’ system. The constitutional recognition of community entrenches communalism, inhibits the development of a unified Mauritian identity, and sits uneasily with modern democratic values.

The first and most fundamental question remains: do we accept that Mauritius is a society composed of distinct communities, and that those communities are entitled to see themselves reflected, in some measure, in the National Assembly? If the answer to that question is no, then the conclusion follows logically: the communal corrective is a mischief and should be removed entirely.

If, however, the answer is yes, then a second and more difficult set of questions arises. Which communities are to be recognised? How are they to be defined? On what basis are their proportions to be determined? And who decides whether a particular candidate belongs to one community rather than another?

The Best loser system answers these questions in a contained manner. It recognises 4 communities and relies on the fixed historical census of 1972 and communal self-declaration. It allows for a maximum of 4 additional seats on communal grounds. Four additional seats in an Assembly of 60+ may well be seen as a modest corrective.

Before deciding whether there is an alternative to it, it is right to ask what communal representation seeks to achieve? If it is intended to protect communal interests, that claim does not actually withstand closer scrutiny. Parliament operates on majority rule. Two, three or more members may not necessarily be able to block legislation. In any case, discriminatory legislation is prohibited by the Constitution. Communal representation cannot meaningfully affect the allocation of public resources (including access to employment opportunities), which must, in any event, be distributed fairly and without discrimination. There are already both ex ante and ex post safeguards here. These safeguards could no doubt be strengthened. There can be more robust equal opportunities legislation, more fairness and transparency in public decision-making, and better access to remedies, including the right of access to, or freedom of, information.

If, on the other hand, the point of communal representation is that it serves a more intangible purpose, a certain symbolism as well as providing reassurance, visibility, a sense of belonging, and a right to political participation, then it must be acknowledged that communal representation is constitutionally relevant.

But does constitutional relevance mean translation into express terms in the Constitution? It may well be if it is to be enforceable. However, this means a further consecration of community-based representation.

This brings us to the critical point. If 4 additional seats are said to be constitutionally objectionable because they entrench communalism, the answer cannot coherently or logically be to increase that number. More than 4 (which I will call the ‘4+ Model’) will only deepen the very problem it purports to solve. It would open the door to further claims, further classifications, and further competition along communal lines, with the difficult questions that courts have found difficult to resolve in Carrimkhan v Tin How Lew Chin (2000) or Narain v The Electoral Commissioner (2005), including on how to resolve the tensions between community classification and individual determination.

It would not matter whether the 4+ Model is one based on proportional representation or another. In fact, one well established critique of proportional representation in constitutional theory is the risk, in multi-ethnic societies, of an increase in ethnically based political parties.

This is why the Mauritius Labour Party was completely against proportional representation in 1965. Our lived history has shown that no communal party has managed any real breakthrough.

Now, there may be hybrid or intermediate models, which seek to enhance inclusiveness without explicit communal classification. A party list is one example, but these offer less transparency or democratic accountability, and in turn, no enforceability.

The choice, therefore, is not quite between 4 (best losers) seats or 4+ seats on communal considerations (through proportional representation or any other model). The choice must bebetween 4 or zero seats.

Zero seat allocated on communal considerations may well be the ideal in a society that has reached a level of institutional strength and social trust sufficient to dispense with communal accommodation altogether. If the consensus is that we, as a nation, are not there yet, then the more modest and contained the accommodation must be the better course. This would mean not opening to the 4+ Model, but to remain on the present model, the best loser corrective. At the same time, we must continue to work to strengthen institutions, political culture, and legal and constitutional safeguards so that communal considerations recede naturally.

Constitutional humility

Ultimately, for present purposes, the central point to retain is that electoral reform is not, as we may be led to believe, the panacea to all ills or all mischiefs. On the contrary, it requires careful and rigorous study, failing which it may open up a Pandora’s box. This article is thus not an argument against electoral reform per se, but an argument for constitutional humility.

Professor Stanley de Smith, our pre-independence Constitutional Commissioner, who had himself fathered several constitutions across the Commonwealth, had initially expressed strong reservations about the best loser corrective when our then leaders proposed and adopted it.

Decades later, when his wife, Professor Barbara de Smith, invited me to Cambridge University to examine his personal papers, she recounted how, in time, her husband had revised his views and came to regard the Mauritian constitutional model as a success. He was, she said, so proud of what Mauritius had become that he asked that his ashes be laid on our island upon his passing.

Constitutional humility lies not in bold redesigns, but in the patient refinement of hard-fought balances that have proved their worth over time.