His case for constituency elections puts a familiar frustration at the centre of the reform debate: voters should be able to identify who represents them, judge their work and replace them. The harder question is how much changing the ballot can achieve.
South Africa can change the relationship between voters and their representatives without abandoning proportional representation. The Constitution leaves room for a different electoral design, provided it produces a broadly proportional result. That makes the question raised by Michael Louis practical: how should voters gain more influence over the individuals who exercise power in their name?
In conversation with Truth Report, Louis argues for constituencies in which candidates must compete on their own credibility, alongside the appeal of their parties. He sees that as a way to attract stronger public representatives and make accountability more personal. His broader ambition is a multiparty government organised around competence and delivery.
A voter should know whom to approach, what that representative has done and how to withdraw support. But Louis’s argument also joins together several distinct decisions: choosing legislators, forming a coalition and selecting the president. Each has its own rules, and success in one does not settle the others.
Making representation personal
Louis’s central proposition is that a recognisable candidate changes the terms of political competition. If one party puts forward an able, trusted person, its rivals must answer with candidates of comparable standing. In his view, the party emblem should carry less of the burden of winning support.
That is a plausible argument about incentives, rather than an established guarantee of better government. A constituency can give a representative an identifiable public to answer to. Whether voters can actually choose between individuals also depends on nomination rules and the ballot itself. Drawing smaller electoral boundaries does not, on its own, determine either.
The legal starting point matters. Section 19(3)(b) protects every adult citizen’s right to stand for public office and, if elected, hold office. That right should be distinguished from a promise that any particular electoral model will produce better leaders.
The Electoral Reform Consultation Panel’s August 2025 Original Report proposed two routes towards closer representation: 300 seats in smaller, multi-member constituencies with 100 compensatory seats, or 200 single-member constituencies with 200 compensatory seats. The compensatory component is intended to restore overall proportionality, so that constituency results do not simply determine each party’s total strength.
These were proposals supported by four named panellists, not a unanimous settlement or an enacted system. The separate Majority or Alternative Report sets out a wider range of options, including changes within the existing regional framework.
The limits of changing the ballot
Voters can already punish parties collectively. The 2024 national election left the ANC with 159 of the National Assembly’s 400 seats, below an outright majority. Accountability through party choice can have substantial consequences.
The Majority or Alternative Report also cautions that an electoral system cannot guarantee accountability. It distinguishes between holding parties collectively responsible and strengthening the relationship with individual representatives. Its discussion of supplementary measures recognises that failures can arise from representatives, parties, legislatures and the executive, as well as the electoral rules.
That is the limit of the business analogy running through Louis’s argument. Recruiting impressive individuals may improve a team. Governing also requires agreement on policy, budgets and responsibility when decisions fail. A coalition of capable people can still disagree about what should be delivered and who should pay for it.
Louis’s preference for cooperation is therefore a political proposition that must be tested against governing arrangements. The constitutional duty already exists: Cabinet members are collectively and individually accountable to Parliament. Electoral reform should help that duty work in practice.
An outsider president still needs Parliament
Louis also entertains the possibility of an outsider helping to lead a broad multiparty government. It is a proposal about leadership, rather than evidence that parties have agreed to such an arrangement.
Under section 86, the National Assembly elects the president from among its own members. An outsider to party leadership would still need to become an eligible Assembly member before being elected president. Public stature alone supplies no alternative route.
Nor does a change of party leader automatically end a state president’s term. Section 88 links that term to a vacancy or the assumption of office by the next elected president. Party succession and presidential succession are separate processes.
There is a limited opening for outside expertise in Cabinet: section 91 allows the president to select no more than two ministers from outside the Assembly. That exception should not be confused with the rules for electing the president.
The work before 2029
Louis wants reform to give voters a more direct connection to public power. Parliament’s 29 August 2026 account of its reform deliberations shows why that ambition requires careful implementation. Committees agreed to develop a roadmap; the meeting did not select a system or settle whether broader reform would take effect in 2029.
The same statement records advice that the legislation establishing the panel does not expressly require wholesale reform by that election. More extensive changes could require new boundaries, revised information systems, testing, training and voter education.
Those are substantial tasks. They also make the demand for a clear public decision more urgent: citizens need to know which relationship Parliament intends to change and how it will judge whether the change works.
Louis’s most useful challenge is to insist that representation should be tangible. Voters should be able to identify a person, assess a record and exercise a meaningful choice. The test of reform will be whether it makes those actions easier, and whether the people elected use their authority to improve the public services on which daily life depends.