Municipal performance management in South Africa rewards paper compliance more than real performance
Introduction
South African municipalities are told, in some detail, how to manage performance. A council adopts an integrated development plan (IDP), the mayor approves a service delivery and budget implementation plan (SDBIP), the municipal manager and the managers reporting to that office sign performance agreements, and at the end of the year an annual report goes to the Auditor-General. On paper this is a tidy chain, and each link refers back to the one before it. Yet the National Treasury classified 64% of municipalities as financially distressed in December 2022 (McKenzie and Marx, 2024), and protest over services has been a regular feature of local politics since the early years of democracy (Masiya, Davids and Mangai, 2021). Something is not connecting.
This paper argues that the chain is working better as a documentation system than as a management system. What it reliably produces are aligned plans, signed agreements and audit opinions. What it produces far less reliably is a municipal official who knows what is expected, a manager who suffers when targets are missed, or a resident who can tell whether services got better (Khunoethe, Reddy and Mthuli, 2021; Bester, 2024). The argument is not that the framework is useless. Audit results in particular do appear to matter to voters (Berliner and Wehner, 2022). The point is narrower, namely that the thing being checked, which is mostly compliance, sits some distance from the thing citizens care about, which is performance.
The paper first sets out what the statutory chain asks for, then looks at where alignment between plans, indicators and individual targets breaks down (Ndamase, Mafunisa and Sethu, 2025). It turns to section 57 performance agreements and the question of consequences (Miya and Matyana, 2024), and after that to audit outcomes, which are the hardest external check the system has (Temoso et al., 2025). It closes with a view on what a more honest measurement of performance would need to include.
What the statutory chain asks for
The Municipal Systems Act 32 of 2000 requires each council to adopt an IDP with a vision, development priorities, a financial plan and key performance indicators (s 25 and s 26). It then requires a performance management system that matches the IDP’s priorities, indicators and targets, promotes a performance culture and works as an early warning of under-performance (s 38 and s 41). Section 57 adds the individual layer: the municipal manager and managers directly accountable to that office hold their posts under a written contract and a separate annual performance agreement, with objectives that are measurable and based on the IDP’s indicators. The Municipal Finance Management Act 56 of 2003 contributes the SDBIP, a detailed plan with monthly projections and quarterly service delivery targets. Finally the annual report, with its performance report and the audit report, has to be tabled and made public (Municipal Systems Act, s 46). Researchers who test this design against practice, such as Khunoethe, Reddy and Mthuli (2021) in Msunduzi and Ndamase, Mafunisa and Sethu (2025) in King Sabata Dalindyebo, start from exactly this sequence of documents.
Read together, these provisions describe a cascade from vision to budget to individual accountability and back out to the public. Studies of how it works tend to find the cascade thinner at the bottom than at the top. In an interview study of KwaDukuza Local Municipality, Miya and Matyana (2024) note that performance management and development was for years confined to section 56 and 57 employees, and that the rollout to staff below that level, which came under the 2021 Municipal Staff Regulations, was still new and poorly understood. Ten interviews is a small base, and one municipality is one municipality (Miya and Matyana, 2024). Still, the study suggests that for most of the workforce the system was, until very recently, something that happened to other people.
There is also the question of how much regulation is too much. Masiya, Davids and Mangai (2021) interviewed key stakeholders and arrived at eleven interrelated factors behind weak official performance, and excessive regulation sits on that list next to weak management systems and a missing culture of excellence. I read that as a warning against treating more prescribed steps as the cure. When the prescribed steps are the thing being counted, officials have every reason to complete them and move on, which is roughly what the symbolic compliance reported by Ndamase, Mafunisa and Sethu (2025) looks like from the inside.
Where the chain breaks: alignment on paper
The best evidence on whether the links actually connect comes from case studies, and they point the same way. Khunoethe, Reddy and Mthuli (2021) examined the Msunduzi Municipality in KwaZulu-Natal using municipal documents and interviews. They found that the key performance areas of top managers did not match the goals in the IDP, so the individual targets were disconnected from the strategic plan they were meant to serve. On top of that, the indicators focused mostly on internal municipal processes instead of outcomes for residents, a few were too vague to measure reliably, and the leadership could not agree on the municipality’s five top strategic objectives. That last finding is worth sitting with. If senior people cannot name the priorities, an agreement that is supposed to be based on them can only be loosely tied to them (Khunoethe, Reddy and Mthuli, 2021; Masiya, Davids and Mangai, 2021).
Ndamase, Mafunisa and Sethu (2025) found something similar in King Sabata Dalindyebo Local Municipality in the Eastern Cape, using semi-structured interviews with municipal officials. Scorecards, monitoring mechanisms and key performance indicators were all in place, and the performance management system had established links between monitoring and evaluation practice and IDP goals. According to the authors, though, those links were often reduced to symbolic exercises with no evidence-based resource allocation or decision-making behind them. Political interference, inadequate data systems and capacity constraints were blamed. Put plainly, the structure existed and the behaviour it was meant to produce did not follow, a pattern that Masiya, Davids and Mangai (2021) also describe when they list weak management systems and weak oversight among the drivers of poor performance.
Bester (2024) approached the problem from the other end. Working with 36 municipal managers from dysfunctional municipalities, selected purposively, he reported that performance was not evaluated satisfactorily, that service backlogs persisted, and that neither the municipalities nor the communities assessed what deficient services did to residents. A sample like that is not representative of all municipalities, and it was chosen precisely because the municipalities were struggling. But the finding fits with the case studies. Plans are written and indicators are set, while the step of asking whether the services actually changed seems to go missing (Khunoethe, Reddy and Mthuli, 2021; Ndamase, Mafunisa and Sethu, 2025).
These studies differ in method and in province, so I would not push them too far. None of them can say how common the pattern is across the country, and two of the three rest on a single municipality each (Khunoethe, Reddy and Mthuli, 2021; Ndamase, Mafunisa and Sethu, 2025). What they can do is show that alignment is not automatic, even where every document exists, and that checking whether a plan refers to another plan is a very different task from checking whether a service improved (Bester, 2024).
Section 57 agreements and the problem of consequences
Performance agreements are meant to be the sharp end of the system, since they attach objectives to named people. The Act also expects them to set out the consequences of substandard performance (Municipal Systems Act, s 57). The evidence on whether consequences follow is thin, which is itself informative. Miya and Matyana (2024) report that participants described leaders and top management as not committed to sustaining a performance culture, that some staff saw performance management as punitive, and that others treated it as a box-ticking exercise with little effect. The authors add that earlier studies had often described the system as a compliance exercise instead of a tool for improvement, and they conclude that clearly communicated consequences and proper support are needed before cascading can progress (Miya and Matyana, 2024; Masiya, Davids and Mangai, 2021).
Why would a signed agreement lose its force? One reading is that the targets inside it are only as good as the plans behind them. If the IDP is vague and the indicators measure internal processes, as in Msunduzi, then a manager can meet every target and still deliver very little that residents would notice (Khunoethe, Reddy and Mthuli, 2021). A second reading is about the environment. Masiya, Davids and Mangai (2021) point to weak oversight, political interference and a tense bureaucratic work environment as part of why officials underperform, and they recommend non-partisan appointments. I will leave the politics of that to other work. For the purposes of performance management, the narrower implication is that an agreement is hard to enforce when the person who would enforce it is operating under the same pressures (Ndamase, Mafunisa and Sethu, 2025).
This whole thing, then, comes down to incentives. Signing an agreement on time is easy to verify and easy to defend. Showing that a manager’s actions improved water reliability or reduced a backlog is harder, takes longer, and depends on data that, according to Ndamase, Mafunisa and Sethu (2025), municipalities often lack. A system will tend to reward whatever it can measure cheaply. In the current design, that seems to be timely documents, which suggests the section 57 mechanism may be doing less to change behaviour than its architects hoped (Miya and Matyana, 2024; Bester, 2024).
Audit outcomes: the hardest check, and what it actually measures
The Auditor-General’s annual audit is the one part of the chain that sits outside the municipality, and it appears to have real effects. Berliner and Wehner (2022) used by-elections triggered by councillors’ deaths and compared those held before and after each year’s audit results were announced. They found that timely audit information moved the vote share of the governing party, with voters rewarding good results and punishing poor ones by roughly five percentage points. That is a meaningful amount, and it suggests that audit outcomes reach citizens in some form and are treated as a signal of how well a council is run. Even so, a signal is only as informative as what sits inside it (Berliner and Wehner, 2022; Temoso et al., 2025).
The difficulty is what the signal contains. Mosimege and Masiya (2022) look at performance reporting rather than finances, although their case is the national Department of Basic Education and not a municipality. Using audit reports and strategic plans from 2010/11 to 2014/15, they found shortcomings in how performance information was managed, and they argue that senior managers needed a better grasp of their oversight responsibilities for pre-determined objectives. Read alongside the municipal case studies, this suggests that the weak point is the quality of the performance information itself, and an audit of that information can only be as meaningful as the objectives it starts from (Mosimege and Masiya, 2022; Khunoethe, Reddy and Mthuli, 2021).
Temoso et al. (2025) then give a useful check on how far audit outcomes track real efficiency. Using a data envelopment analysis of South African municipalities for 2017 to 2022, followed by a fractional regression, they report that audit outcomes had no direct effect on municipal efficiency. The interaction with municipality type told a more textured story, since metropolitan municipalities receiving disclaimers or qualified opinions saw clear efficiency declines in the pandemic years. Even so, the headline result points toward a loose link between the audit opinion and how well resources are turned into services. I would not read it as saying audits are pointless. It does suggest that a clean opinion and a well-run municipality are separate things that sometimes happen together (Temoso et al., 2025; Berliner and Wehner, 2022).
The same loose link shows up when the state intervenes. Khaile (2023) studied municipalities in Limpopo, North West and the Free State placed under section 139 intervention and used audit outcomes as a stand-in for financial accountability. A year on, Limpopo’s municipalities had improved, those in North West and the Free State had not, and even in Limpopo none reached a clean audit. McKenzie and Marx (2024), working from audit action plans and a census of provincial treasuries, the National Treasury and the Department of Cooperative Governance and Traditional Affairs, found that turnaround strategies built on internal controls, revenue management, and ethics and fraud prevention were associated with better financial viability. This is useful, because it shows where audit-driven improvement seems to take hold. It is financial control, which is the area audit reports are built to examine (Khaile, 2023; McKenzie and Marx, 2024).
Measuring what residents would recognise
None of this means the framework should be scrapped. The statutory chain gives the system a shared vocabulary, and the audit gives it an outside check that voters evidently notice (Berliner and Wehner, 2022). What the evidence suggests is that the system needs to measure a different set of things, and to attach consequences to them. The first step would be indicators that describe outcomes for residents, such as hours of water supply or response times to faults, instead of internal process milestones. The case studies suggest this is a gap in practice more than in law, since the Act already calls for indicators covering outcomes and impact (Municipal Systems Act, s 41), and Msunduzi’s indicators still leaned toward internal processes (Khunoethe, Reddy and Mthuli, 2021; Ndamase, Mafunisa and Sethu, 2025).
The second step is to publish performance in a form residents can compare. If an audit opinion moves votes by about five points, a comparable and credible measure of service performance might move them further, and it would give councillors a reason to care about the indicators in the IDP (Berliner and Wehner, 2022; Bester, 2024). That is a hypothesis and not a finding, to be fair. Whether the data capacity exists is also a fair worry. Ndamase, Mafunisa and Sethu (2025) found inadequate data systems in their case municipality, and Miya and Matyana (2024) found no electronic system at all in theirs. That points to a sequence, with data systems first and public comparison later.
The third step is to read efficiency alongside the audit opinion, so that a municipality with a clean audit and poor value for money is not treated as a success story (Temoso et al., 2025; McKenzie and Marx, 2024). None of these proposals is expensive in principle. All of them do depend on someone being willing to publish figures that are not flattering, which is a risk to any council and may be part of why the current arrangement has lasted (Masiya, Davids and Mangai, 2021).
Conclusion
The performance management chain in South African local government is complete on paper and uneven in practice. The studies reviewed here show indicators that do not match the IDP, scorecards used symbolically, agreements that carry few visible consequences and audit opinions that capture financial compliance far better than service delivery (Khunoethe, Reddy and Mthuli, 2021; Miya and Matyana, 2024; Temoso et al., 2025). That does not make compliance worthless. A municipality that cannot keep clean records is unlikely to deliver much else, and voters do punish poor audits (Berliner and Wehner, 2022; McKenzie and Marx, 2024).
My position is that the system has been tuned to what can be verified quickly, and that this helps explain why more rules have not produced better services. The evidence base is mostly qualitative, drawn from a handful of municipalities, and the one large quantitative study covers only six years (Temoso et al., 2025; Bester, 2024). So the conclusions are tentative. Even so, the direction seems clear enough: measure outcomes residents would recognise, make them public, and read them next to the audit opinion instead of behind it (Mosimege and Masiya, 2022; Khaile, 2023).
Reference list
Berliner, D. and Wehner, J. (2022) Audits for accountability: evidence from municipal by-elections in South Africa. The Journal of Politics, 84(3), pp. 1581-1594. doi:https://doi.org/10.1086/716951
Bester, J. (2024) Making dysfunctional municipalities functional: towards a framework for improving municipal service delivery performance in South African municipalities. Commonwealth Journal of Local Governance, (29), pp. 59-78. doi:https://doi.org/10.5130/cjlg.vi29.8702
Khaile, T. (2023) The effectiveness of Section 139 interventions in strengthening municipal accountability in South Africa. Journal of Local Government Research and Innovation, 4, article a146. doi:https://doi.org/10.4102/jolgri.v4i0.146
Khunoethe, H., Reddy, P.S. and Mthuli, S.A. (2021) Performance management and the integrated development plan of the Msunduzi Municipality in South Africa. NISPAcee Journal of Public Administration and Policy, 14(2), pp. 161-187. doi:https://doi.org/10.2478/nispa-2021-0019
Masiya, T., Davids, Y.D. and Mangai, M.S. (2021) Factors affecting the performance of South African municipal officials: stakeholders’ perspectives. Commonwealth Journal of Local Governance, (25), pp. 97-115. doi:https://doi.org/10.5130/cjlg.vi25.7701
McKenzie, M. and Marx, B. (2024) Turnaround for municipal distress resolution: an audit outcomes approach. South African Journal of Accounting Research, 38(2), pp. 174-208. doi:https://doi.org/10.1080/10291954.2023.2294536
Miya, A. and Matyana, M. (2024) Critical analysis of cascading for performance management system to South African local municipality. Journal of Public Administration, Finance and Law, (32), pp. 295-306. doi:https://doi.org/10.47743/jopafl-2024-32-22
Mosimege, S.B. and Masiya, T. (2022) Assessing audit outcomes of pre-determined objectives in a public institution (South Africa). International Journal of Productivity and Performance Management, 71(6), pp. 2296-2315. doi:https://doi.org/10.1108/IJPPM-02-2021-0091
Ndamase, M., Mafunisa, M.J. and Sethu, J. (2025) Assessing the alignment between integrated development plan goals and monitoring and evaluation indicators: a case of King Sabata Dalindyebo Local Municipality. The African Journal of Governance and Development, 14(2), pp. 41-60. doi:https://doi.org/10.36369/2616-9045/2025/v14i2a3
Temoso, O., Tran, C.T.D., Tran, J. and Myeki, L. (2025) South African local government efficiency and the effect of audit outcome. Politics & Policy, 53(6), article e70098. doi:https://doi.org/10.1111/polp.70098
Legislation and case law
Local Government: Municipal Finance Management Act 56 of 2003. Available at: https://www.gov.za/sites/default/files/gcis_document/201409/a56-03.pdf (Accessed: 18 October 2026).
Local Government: Municipal Systems Act 32 of 2000. Available at: https://www.gov.za/sites/default/files/gcis_document/201409/a32-000.pdf (Accessed: 18 October 2026).