Navigating the Ideological Divide: The Litmus Test for Fragile GNU

Prof Dirk Kotzé

South Africa’s second Government of National Unity (GNU) reached its first anniversary this week. Comparisons with the GNU of 1994 were initially made, but the ensuing debates made it clear that they are not comparable. 

The first point of debate was whether the new multiparty government was indeed a GNU or a coalition government. For some, the distinction was merely a matter of semantics, while for others, it represented a political stance on black unity in government.

The fact that the uMkhonto weSizwe party (MKP) and Economic Freedom Fighters (EFF) were not included in the new government disqualified it as a national unity government. In the meantime, this debate has petered out, mainly because the ANC and DA continue to use the name GNU.

The GNU’s formal foundation is very limited in nature: in essence, it is the Statement of Intent first signed by the ANC and DA, and later by the other eight parties. The Statement included ten Foundational Principles, the Basic Minimum Programme of (nine) Priorities and the GNU’s Modalities or main procedural arrangements. The programme of priorities came the closest to an agreement about policy matters, though they were formulated in very general terms.

 Compared with new coalition governments in many parts of the world who first negotiate a coalition agreement over an extended period, a combination of sections 51 and 86 in the South African constitution determines that a new president must be elected by Parliament not more than 14 days after the results of the election were announced.

That sets the maximum time available to establish a coalition of at least 50%+1 in parliament that is required to elect a president. The fact that the election of the president is linked to a parliamentary majority (and therefore in some cases a coalition) is because the president is not directly elected. The coalition also does not allow for the outgoing cabinet to continue until a new one is formed by the president but is dissolved when the new president is inaugurated (section 94) five days after the presidential election.

The impact of these constitutional arrangements are lessons already learnt from the GNU’s first year of existence. Given the fact that coalition governments appear to be the norm in South Africa for the foreseeable future, constitutional re-engineering might be required, because the GNU’s early history demonstrated that the Constitution was not designed to include a coalition government.

A potential complication for a coalition government – especially for one in a stalemate situation – is the Constitution’s section 50(1)(b) which states that an early general election cannot be held within three years after the previous election. It means that no early election can be held before May 2027.

Coalition governments sometimes collapse, like the German one not long ago, and then an early election is held to form a new government. If it were to happen in South Africa before 2027, such an option would not exist. Its desirability in South Africa under new conditions should be considered.

A counterintuitive consequence of the GNU’s emergence is that it enhanced the prominence of Parliament. In the past, the ANC government (cabinet) dominated Parliament, especially when the opposition was small. The ANC also did not accommodate parliamentary backbenchers who could ask difficult questions. With the GNU, these dynamics are changing. Portfolio committees are now chaired by members of the GNU parties and even other small parties such as Rise Mzansi, BOSA and ActionSA. DA and other GNU parliamentarians play often the role of critical backbenchers.

The official opposition (MKP) and the EFF are often vocal in debates but not necessarily an effective opposition that can enhance public accountability by the GNU executive. In this regard, the oversized 70% coalition government suffocates the parliamentary opposition. That might be one of the considerations why GNUs are normally only a one-term government arrangement.

It is noteworthy that opinion surveys done by the Social Research Foundation suggest that the bigger parties in the GNU (except for the ANC) have increased their support since the May 2024 elections, while the opposition parties have lost support, notably the MKP and EFF. It suggests that the downside of the GNU already mentioned, as well as the policy differences, the GNU as a concept, is not discredited in the public mind.

Despite the absence of a coalition policy agreement – or maybe because of it – policy matters have become very prominent during the GNU lifespan. South African politics are dominated by the socio-economic issues of increasing unemployment, crime, illegal immigration and transformation policies, as well as the governance issues of service delivery and corruption.

In the context of the GNU, the policy issues that became prominent have been school education (BELA Act), health care (NHI Act), fiscal policies (Budget), property rights (Expropriation Act) and international relations. In the areas of home affairs, agriculture and trade & industry a process of convergence is seen.

While the conventional approach for coalitions is to negotiate an agreement on policy matters, largely as a consequence of the 14-day limitation on the formation of a coalition, the alternative is a piecemeal approach to dealing with policy issues one by one as they emerge. Some of them, like the BELA Act, went to the GNU’s “clearing house” or dispute resolution mechanism, chaired by Deputy President Mashatile, and thereafter to the National Economic Development and Labour Council (Nedlac).

The Budget issue was resolved by different negotiation and consultation processes, mainly by the ANC and DA, while other policy matters, like the BELA and NHI Acts as well as the Budget, were taken to court. The result so far is therefore an incremental development of a policy framework in the absence of a comprehensive agreement.

The 2025-26 Budget stands out as a major test for the GNU. The fact that after three attempts the parties could find each other is an indication of how important it is for the ANC and DA that the GNU does succeed. The ANC has about three other options for partners if the relations with the DA falter, but none of them is an attractive one. 

Remarkably, President Trump’s threats aimed at South Africa and his genocide/farm murders/land grab accusations united the GNU partners against these misrepresentations. Another serious test for the GNU will be the 2026 local government elections when all the parties will campaign against each other. Possible leadership changes in the DA (at their 2026 Federal Congress) or the ANC (2027 National Conference) will also be a litmus test. If the GNU prevails despite leadership changes, then it implies a sound institutionalization beyond dependence only on personalities.

* Prof Dirk Kotzé, Department of Political Sciences, Unisa

** The views expressed do not necessarily reflect the views of IOL, Independent Media or The African.