A week ago, The Common Sense published an editorial proposing that Johannesburg should be broken up into smaller municipalities. The argument was that a single municipal government serving almost six million people across an enormous and economically diverse urban area has become too distant from the residents it is supposed to serve. Since then, there has been useful feedback about how such a proposal might work in practice, particularly around the question of how Johannesburg could be divided, while retaining common infrastructure and services, where metropolitan scale remains useful. What follows is a draft proposal for how such a system could work (and a warning that something similar will happen whether the politicians get on board or not).

Johannesburg could be divided into several independent municipalities, each with its own elected council, mayor, administration, budget and tax base. Those municipalities would be directly accountable to the people living within their boundaries, and would control the overwhelming majority of ordinary municipal functions. Water, parts of the electricity network, major transport systems, landfill infrastructure, and selected administrative systems could continue to operate across municipal boundaries through jointly owned utilities, and service agreements between the new municipalities.

The existing seven administrative regions provide a useful starting point. Johannesburg already administers itself through Regions A to G, which broadly correspond with recognisable parts of the city. Region A includes Midrand, Fourways, Diepsloot, and Ivory Park. Region B includes Randburg, Rosebank, Northcliff, and surrounding areas. Region C is centred broadly on Roodepoort. Region D consists largely of Soweto. Region E includes Sandton, Alexandra, and much of northeastern Johannesburg. Region F includes the inner city and surrounding southern and eastern areas, while Region G extends into the deep south through areas including Lenasia, Ennerdale, and Orange Farm.

The proposal is not that these regions be retained – rather that the concept they speak to be pursued into a new set of municipalities where the interests of the residents align into viable stand-alone entities that are either self-financing or cross-subsidised by the national government where the tax base is not sufficient to support a viable municipal administration.

Plausible municipalities

A plausible structure could contain between five and 10 municipalities. Johannesburg North could incorporate much of present Region A around Midrand and Fourways. Johannesburg West could combine much of Regions B and C around Randburg, Northcliff, and Roodepoort. Soweto could form another municipality based substantially on Region D. Johannesburg East could incorporate Rosebank and Sandton and adjoining northeastern areas. Regions F and G could form a Central and South municipality or could be reorganised into separate Johannesburg Central and Johannesburg South municipalities, if the fiscal and administrative analysis supports that arrangement.

The final number should emerge from detailed modelling of the areas concerned. Each new municipality would require a sufficient revenue base, manageable administrative geography, infrastructure alignment, and a population large enough to sustain a substantial municipal administration. Property values, rates revenue, commercial and industrial activity, service obligations, infrastructure assets, transport connections, settlement patterns, and future capital requirements would all need to be assessed before final boundaries were drawn.

The central test is viability. Each municipality must be capable of functioning as an independent government. Some would inevitably be considerably wealthier than others, and could even stand entirely on their self-generated revenue. That variation already exists across South African municipalities and can be accommodated within the country's existing system of intergovernmental finance.

Support

Poorer municipalities would receive support through the national fiscal system. The local government equitable share, conditional grants, and other national and provincial transfers, already distribute resources towards municipalities with weaker own revenue bases and greater service needs. This makes sense and alleviating institutional poverty and underdeveloped infrastructure is not a municipal function. Rather, national government revenues should be assigned to lifting such poor districts to the point where they become viable. Wealthier suburbs should pay their own way.

South African local government already provides several institutional forms that are relevant to such a proposal. Municipalities fall into three broad categories. A Category A municipality is a metropolitan municipality exercising municipal authority across its own area. Johannesburg, Cape Town, Tshwane, and Ekurhuleni are examples. A Category B municipality is a local municipality that shares municipal responsibilities with a Category C district municipality. The Category C municipality covers several Category B municipalities and performs functions assigned to the wider district.

Johannesburg currently operates as a Category A metropolitan municipality. A practical route would be to divide the existing metro into several smaller Category A municipalities. Each new municipality would then exercise municipal executive and legislative authority within its own boundaries, elect its own council, approve its own budget, levy its own rates, and appoint its own administration.

The successor municipalities would therefore be governments in their own right. Johannesburg West would have its own council and mayor. Johannesburg East would have its own council and mayor. Soweto would have its own council and mayor. Johannesburg North and the Central and South municipalities would operate in the same way. Each would prepare its own budget, determine its own spending priorities, manage its own administration and have its own electorate.

The scale of these municipalities would remain substantial. With Johannesburg's population being between five and six million people, a division into five or 10 municipalities would generally produce councils governing populations ranging from several hundred thousand to more than one million residents. These would remain major urban governments with large budgets, substantial professional administrations, and significant infrastructure responsibilities. Their scale would be considerably closer to the communities they govern than the present metropolitan structure.

That change would make political accountability much clearer. In short, local government should be local, and local politicians should be accountable to a relatively small voter base.

The structure would also allow different councils to make different policy choices. One municipality might place greater emphasis on infrastructure maintenance and capital investment. Another might focus more heavily on keeping rates increases low. A council could streamline planning and development approvals in an attempt to attract investment. Another could choose a different spending mix. The fiscal and service consequences of those decisions would become visible within each municipality.

Over time, these differences could produce useful competition between governments. Businesses considering new investment would be able to compare municipal rates, approval times, infrastructure quality, and service reliability. Residents would be able to compare roads, parks, refuse collection, and local facilities. Municipal performance would become easier to observe because each council would govern a smaller, and more coherent, territory.

Infrastructure

The key infrastructure systems could also continue operating across municipal boundaries. Water networks, strategic electricity infrastructure, landfill facilities, and public transport systems can serve several municipalities through common institutions.

With respect to Joburg Water and City Power, it would be best for each municipality to purchase bulk supplies directly from the utility. The utility can enforce accountability by holding each municipality accountable for its usage and debts.

Roads would generally become a local responsibility. Major cross-boundary routes or specialist functions could be managed jointly where this produced operational advantages, but routine road maintenance would sit firmly with the municipality whose voters use those roads.

Waste collection would similarly become a local municipal function. Each council would provide, or contract, household refuse collection within its jurisdiction and carry responsibility for the quality of that service. Large landfill sites, transfer stations and processing infrastructure could continue to serve several municipalities through shared ownership and operating agreements.

Public transport would require some cooperation between the successor municipalities. The simplest would be just to outsource most of it to the taxi industry which provides far and away the most efficient means of moving people around the city. In time, integrated high-speed rail spurs and the like could be driven by the national and provincial government in cooperation with whichever of the local municipalities is apt.

Selected systems

Municipal cooperation could also extend to selected administrative systems. Billing technology, property valuation, specialist engineering services, information technology, and certain procurement functions could initially be shared where common systems reduce costs – although in the current case the waste and corruption in centralised procurement is so vast that breaking it up could be cheaper. Each municipality would still control its own budget, revenue, personnel, and policy decisions. Shared systems would provide services to the municipalities rather than exercise political authority over them.

This approach would also create clearer financial accountability. Each municipality would have its own budget and balance sheet. Revenue collected within the municipality, together with national transfers and service charges, would finance that government's operations. A council that maintained strong revenue collection and controlled expenditure would benefit from that performance. A council that allowed its finances to deteriorate would confront the consequences directly within its own municipality.

The national transfer system would play an important role in supporting municipalities with weaker tax bases. The national government already allocates substantial funding to local governments according to population, poverty, service responsibilities, and fiscal capacity. The creation of a Soweto municipality or a Johannesburg South municipality would therefore take place within an established national system designed to support jurisdictions with limited own revenue.

Shared utilities would require separate financial arrangements. Each participating municipality would contribute towards the cost of common infrastructure according to an agreed formula. Water costs could be linked to consumption and capital requirements. Transport contributions could reflect routes, passenger volumes and infrastructure use. Waste infrastructure could be financed according to volumes or capacity requirements. These arrangements would give each council a clear understanding of what it was paying for common services.

The transition would require detailed work on Johannesburg's existing assets and liabilities. Municipal debt would need to be allocated between successor municipalities and shared utilities. Property, vehicles, infrastructure, pension obligations, employment contracts, and procurement commitments would need to be assigned. Existing capital programmes would need to be separated according to the areas and infrastructure they serve.

What about the pushback? The African National Congress (ANC) will kick back at this proposal given its instinct for centralised control of things. But it should think carefully. It stands to lose all of Johannesburg. Under this proposal, if it plays its cards right, it is guaranteed control of large municipal districts across Gauteng funded out of national coffers. There is an upside for the Democratic Alliance in that it could get on with governing areas of the city that it could not otherwise control.

What would happen if this proposal is ignored?

The reality is that what is being proposed here is already happening, and is going to happen to an increasing degree, regardless of whether the politicians support it or not. As the city government collapses, and the national government on top of it, people are already moving to segment the city into the enclaves this newspaper spends so much time writing about. The middle-class suburbs, business parks, and major retail centres are becoming private estates – private municipal districts in all but name. Private security firms, the taxi industry, and illicit economic actors now enforce law and order, not the national or city police.

In poorer communities, gangs and related actors are supplanting the control the ANC once exercised. The splintering of the ANC into the Economic Freedom Fighters and uMkhonto weSizwe Party only exacerbates this trend. The example of how the illicit mining industry now commands control of parts of the city outskirts is an example of what the non-middle-class enclaves will all become.

As usual, the people of South Africa are already two steps ahead of the politicians. Anyone debating the legality of this proposal should spend some time witnessing how zama-zama enforcers operate on the peri-urban edge of Johannesburg. The fragmentation of massive municipalities is coming. In fact, it is already here. It would be best if politicians moved with this trend so as to ensure that a lawful, ordered system of local government remains relevant in South Africa.