Intergovernmental forums and the district development model coordinate Gauteng’s metros by consultation alone

Introduction

South Africa’s metros are meant to be coordinated by two instruments that sit on top of each other. The Intergovernmental Relations Framework Act 13 of 2005 sets up forums, protocols and a dispute procedure, and the District Development Model, announced in 2019, adds an approach in which the three spheres align their plans and budgets in 52 development spaces, eight of them metros (Republic of South Africa, 2005; Department of Cooperative Governance, no date). On paper that is a lot of machinery. The evidence on whether it actually coordinates anything in the big cities is thinner than the policy documents suggest, and what exists comes mostly from districts and from small samples (Tafeni, 2024; Teti et al., 2024).

This paper argues that both instruments coordinate mainly by consultation. They bring people into rooms and ask them to align, but nothing in them obliges a sphere to give way when interests, budgets or politics point in different directions (Republic of South Africa, 2005). The Gauteng metros show why that matters, because their coalition politics, their late budgets and the water problem in Hammanskraal all involve choices that no forum can settle by itself (Pholoma et al., 2024; Zweni et al., 2024). Hammanskraal is a particularly clear example (Makgopela and Radikonyana, 2023). I take Gauteng as the case because the recent metro studies I could verify cluster there, and I use Eastern Cape district studies only for what they say about the model. The paper goes through the Act, then the model, then Gauteng’s coalitions, then Hammanskraal, and ends with what would give coordination some teeth.

What the Framework Act provides

The Act’s aims are sensible. They include coherent government, effective service delivery and the monitoring of implementation, and each sphere is told to consider the others’ budgets, coordinate actions that affect them and avoid jurisdictional contests (Republic of South Africa, 2005). The structures that carry those duties are consultative by design. The President’s Co-ordinating Council is described as a consultative forum, the Premier’s forum is a consultative forum for a province and its municipalities, and the district forums follow the same pattern (Republic of South Africa, 2005). None of them decides anything on behalf of another sphere. That is not a drafting accident. It probably follows from a Constitution that treats each sphere as distinct, and the forums are where those distinct spheres talk.

Metros are an odd fit. A metro mayor sits on the Premier’s forum, but the Act provides no district forum for metros, because a district forum brings together a district and the local municipalities inside it (Republic of South Africa, 2005). So the largest and most complicated municipalities have no forum of their own, and share a table with the province’s MEC and the other mayors. Drewes et al. (2025) describe regional governance, the layer that includes provinces, districts and metros, as playing more of an advisory and compliance role in spatial planning than a deciding one. That points the same way as the Act: the forum advises, and whichever sphere holds the statutory function decides.

The binding parts of the Act are optional or late. Implementation protocols are available where a service depends on organs of state in different governments, and they must set out roles, indicators, resources and dispute procedures, but the Act says the organs may use them, not that they must (Republic of South Africa, 2005). The dispute chapter then requires every reasonable effort to settle a dispute before anyone goes to court, lets the Minister or the MEC convene the parties, and bars court action unless a dispute has been formally declared and those efforts have failed (Republic of South Africa, 2005). That is a good rule for avoiding litigation. It is a weak rule for producing delivery, since a stalled project is rarely a formally declared dispute. Pietersen (2025) adds that scholars and government’s own evaluation reports have found weaknesses and limited success in intergovernmental relations since the start of democracy, well before the present coalition era.

The District Development Model and what its evidence covers

Tafeni (2024) quotes the President’s 2019 complaint about a pattern of operating in silos, which left planning without coherence and made oversight of government programmes difficult, and the model is the answer to that. His own test is the OR Tambo pilot in the Eastern Cape, where all fifteen questionnaires he distributed came back, and the respondents suggest the model has to a large extent improved coordination and integration. He still calls for a structured and legislative coordination framework behind it (Tafeni, 2024). Teti et al. (2024), interviewing five senior managers in the same district, found general awareness of the model but grey areas about how it is meant to work and how results will arrive, especially in coordination and communication inside the model. The same pattern appears one level up. Tafeni and Mngomezulu (2022) report that the Eastern Cape’s provincial MuniMEC has contributed to coordination at municipalities, yet little empirical evidence explains the improvement, and they warn that regulation without a structured plan for implementing it leads to poor service delivery.

Against that, Nel and Minnie (2022) conclude that the model may not succeed because it does not address the core problems of local government, which they list as weak capacity and resources, poor governance and limited coordination between and within governments. They want a candid review of what local government is, what powers it has and how it relates to the other spheres (Nel and Minnie, 2022). Mashapha and Makole (2025), working from literature and documents, take the opposite view, arguing that the model can meet the aims of section 153 of the Constitution and that it carries a change management strategy with monitoring, evaluation and consequence management. The disagreement partly looks like a difference in method. The optimistic findings come from participants in a pilot, or from a reading of what the policy intends, and the sceptical one from an analysis of what the policy leaves untouched. Tafeni’s own call for a legislative framework sits closer to Nel and Minnie than his headline finding suggests.

What matters for this paper is that none of these studies is about a metro. The pilot evidence is from a district, the samples are fifteen and five people (Tafeni, 2024; Teti et al., 2024), and the rest is literature review (Mashapha and Makole, 2025; Nel and Minnie, 2022). A metro is a single municipality covering a whole city, with no local municipalities beneath it, so the district style of lining up local and district plans seems to have less to bite on there. That is an inference from structure, not a finding. Even so, the model’s own list of spaces includes eight metros (Department of Cooperative Governance, no date), and the research has not caught up with that fact, which means the places where the stakes are highest are the places where we know least about whether the model works.

A fair test would need more than impressions. Coordination can be measured, at least roughly, by asking whether the budgets of two spheres actually line up in a given year, whether projects are jointly funded and finished, and whether a metro and a department stop contradicting each other’s plans. The studies above mostly record what participants think. Tafeni (2024) has fifteen respondents, Teti et al. (2024) have five managers, and Tafeni and Mngomezulu (2022) say outright that little empirical evidence explains the improvement they report. Perceptions from inside a coordinating body are worth having, and they are also exactly what a body would be expected to say about itself when it has been asked to coordinate. That is not an accusation, only a reason to want some harder numbers before the model is declared a success in places like Johannesburg or Tshwane.

Coalitions in Gauteng’s metros

Coordination needs a stable counterpart, and Gauteng’s metros have often not been one. Pholoma et al. (2024) studied Tshwane, Ekurhuleni and Johannesburg using secondary sources and thematic analysis, and found that unstable coalitions erode accountable governance, hurt the provision of services, compromise budget formulation and approval, contribute to hung councils and affect municipal human resource practice and fund management. Zweni et al. (2024) point out that the 2021 local elections left nine hung councils in Gauteng, and their financial ratio analysis of the metros under coalitions suggests revenue that is not sustainable, expenditure that is not credible, budgets that do not respond to the needs of the poor and budgets passed late. They also report that coalition parties do not keep to their own agreements (Zweni et al., 2024).

The link to intergovernmental relations is fairly direct. A model that asks three spheres to align plans and budgets cannot do much with a metro whose own budget is late or unsettled (Zweni et al., 2024; Department of Cooperative Governance, no date). Pietersen (2025) takes the optimistic side, noting that coalitions are constitutionally required to govern within the intergovernmental framework, and that effective relations through cooperation, transparency and structures may help to stabilise them. I think the causal arrow probably runs more strongly the other way. The conceptual analysis is useful, but it is conceptual, and the two Gauteng studies both describe instability damaging the very budgets and services that coordination is supposed to protect (Pholoma et al., 2024; Pietersen, 2025). Neither tests whether a Premier’s forum moderates any of it. Both Gauteng papers also rely on secondary data and appear in the same journal in the same year, so they probably lean on overlapping material, which is a reason to hold their agreement a little loosely.

Coalitions are not the whole story either. Pietersen (2025) notes that the weaknesses in intergovernmental relations were already visible when the ANC dominated all three spheres, so coalition instability looks more like an aggravating factor than the origin of the problem. That matters for how we read the Gauteng studies. If the foundations were already consultative and thin, then a hung council in Johannesburg or Tshwane does not so much break a working system as expose one that relied on goodwill between stable partners (Pietersen, 2025; Pholoma et al., 2024). Under a dominant party, informal party channels could patch over gaps that a formal forum left open. Once those channels weaken, the forum is all that is left, and the forum was never built to compel.

Hammanskraal as a coordination test

Makgopela and Radikonyana (2023) offer a case that looks like a coordination problem even though their abstract never uses the phrase. Their qualitative study found that the water in Hammanskraal, within Tshwane, is not fit for human consumption, and that problems run through different components of the supply system. Residents were concerned about the quality of the water the municipality provides, while Tshwane officials downplayed those concerns. Their remedy is a more concerted effort from government, environmental agencies and the public, together with better water supply management by the local government (Makgopela and Radikonyana, 2023).

What does the case show? Mainly a gap between a clear statutory owner and a shared problem. The metro is formally responsible for providing the water, so it is the one residents will look to, yet the fix described in the study spans several actors (Makgopela and Radikonyana, 2023). The Act asks each sphere to coordinate actions that affect the others and lets organs sign protocols where a service depends on more than one of them (Republic of South Africa, 2005). Whether such a protocol or a forum was ever used for Hammanskraal is not something these sources let me say, and I would rather not guess. The point is narrower. A resident whose tap water is unsafe has no dispute to declare, the sphere that could help has no duty to answer to the resident, and the metro is left to carry a problem it cannot solve alone. Forums that consult do nothing about that.

What would give coordination some teeth

The district studies, for all their limits, converge on the same missing piece. Tafeni (2024) asks for a structured and legislative framework, Tafeni and Mngomezulu (2022) warn that regulation without a structured implementation plan fails, and Nel and Minnie (2022) want a rethink of powers and relationships. Teti et al. (2024) come at it from another direction, recommending an open systems approach in which the model’s hub works as an agile team, which is basically a request for structure that is flexible but still there. Mashapha and Makole (2025) likewise end on monitoring and consequence management. Even the supporters of the model, in other words, keep reaching for something that bites. So the first change is to make protocols the normal vehicle for services that cross spheres in a metro, with named roles, money and indicators, which the Act already lists as the content of a protocol (Republic of South Africa, 2005). At present they are available. They should be expected for the functions where Gauteng’s problems cluster, such as bulk water and transport, even if the evidence for exactly which functions is indirect.

The second change is to give metros a table of their own, or at least a standing metro item on the Premier’s forum, since the Act has nothing designed for a one-tier city (Republic of South Africa, 2005; Drewes et al., 2025). The third is to evaluate the model in the metros using measures other than participants’ impressions, because the only favourable evidence so far comes from respondents inside the pilot (Tafeni, 2024). And the fourth sits outside intergovernmental relations altogether. Pholoma et al. (2024) conclude that government and municipalities need legislation and policies to deal with unstable coalitions, and without that, forums will keep meeting a counterpart that cannot commit its budget. More legalism has its own risk, of course, since binding protocols could multiply formal disputes. Still, that seems a better problem to have than the current one.

Conclusion

The Framework Act and the District Development Model are both built to bring the spheres into conversation, and they do that reasonably well. What they do not do is oblige anyone to yield (Republic of South Africa, 2005; Nel and Minnie, 2022). In Gauteng’s metros the choices that matter, such as who passes a budget on time, who answers for unsafe water and whether a coalition holds, sit inside one sphere or between two, and a consultative forum has little to say about any of them (Pholoma et al., 2024; Zweni et al., 2024; Makgopela and Radikonyana, 2023). The evidence for the model is also lopsided, with favourable findings from small samples in a district and scepticism from structural analysis, and next to nothing from the cities it was meant to include (Tafeni, 2024; Mashapha and Makole, 2025). I do not think the answer is to abandon either instrument. Obligations, money and a metro-specific forum need to be attached to them, and the results need to be measured where the stakes are highest, because at the moment the success of the model in the cities is taken on trust.

Reference list

Drewes, J.E., van Aswegen, M. and Brand, A. (2025) ‘Regional governance and spatial transformation in South Africa post-1994’, in du Plessis, D. (ed.) Territorial governance and spatial transformation in post-apartheid South Africa. Cham: Springer Nature Switzerland, pp. 17-43. doi:https://doi.org/10.1007/978-3-031-85489-7_2

Makgopela, P.S. and Radikonyana, P.S. (2023) Drinkable and potable water in Hammanskraal, Tshwane Metropolitan Municipality: a service delivery challenge or a fallacy? Journal of Public Administration, 58(2), pp. 284-299. doi:https://doi.org/10.53973/jopa.2023.58.2.a4

Mashapha, M.C. and Makole, K.R. (2025) District Development Model as a panacea to poor performance in local government municipalities in South Africa. E-Journal of Humanities, Arts and Social Sciences, 6(14), pp. 4023-4033. doi:https://doi.org/10.38159/ehass.202561446

Nel, V. and Minnie, S. (2022) Can the district development model deliver development in South African local governments? Rozwój Regionalny i Polityka Regionalna, (60), pp. 39-53. doi:https://doi.org/10.14746/rrpr.2022.60s.05

Pholoma, M., Lubinga, S., Masiya, T. and Madumo, O.S. (2024) The influence of unstable coalition governments in Gauteng metropolitan municipalities. Journal of Local Government Research and Innovation, 5, article a195. doi:https://doi.org/10.4102/jolgri.v5i0.195

Pietersen, J.M. (2025) Intergovernmental relations in fledgling coalition governments of South Africa. Journal of Public Administration, 60(1-1), pp. 391-407. doi:https://doi.org/10.53973/jopa.2025.60.1.1a9

Tafeni, Z. (2024) Assessment of the District Development Model implementation impact towards improvement of coordination and integration of support in the local government environment, Eastern Cape Province South Africa. Journal of Public Administration, 59(1), pp. 119-129. doi:https://doi.org/10.53973/jopa.2024.59.1.a10

Tafeni, Z. and Mngomezulu, B. (2022) Assessing the effectiveness and efficiency of the provincial MuniMEC forum inter-government relations structures in the Eastern Cape Province, South Africa. Journal of Local Government Research and Innovation, 3, article a42. doi:https://doi.org/10.4102/jolgri.v3i0.42

Teti, M., Nzewi, O.I. and Lungisa, S. (2024) The district development model as a catalyst for improved integrated development planning. Journal of Local Government Research and Innovation, 5, article a153. doi:https://doi.org/10.4102/jolgri.v5i0.153

Zweni, A., Koma, S. and Ndevu, Z. (2024) Coalition effects on financial and service delivery performance in metropolitan municipalities in Gauteng. Journal of Local Government Research and Innovation, 5, article a183. doi:https://doi.org/10.4102/jolgri.v5i0.183

Legislation and case law

Department of Cooperative Governance (no date) District Development Model. Available at: https://www.cogta.gov.za/ddm/ (Accessed: 18 October 2026).

Republic of South Africa (2005) Intergovernmental Relations Framework Act 13 of 2005. Available at: https://www.gov.za/sites/default/files/gcis_document/201409/a13-051.pdf (Accessed: 18 October 2026).