Centralised Decentralisation: The strings we do not see, and what managing Tema has taught me

1. A Conversation in Kigali

The African Academy for Women in Political Leadership convened us in Kigali in the first week of August, 2026. 40 women from 28 countries in Africa and the Caribbean. Ministers, legislators, mayors, party leaders. I was one of two Ghanaians.

We had come to learn particularly from Rwanda’s remarkable trajectory in women’s political leadership, our unique and shared experiences, and to think together about how women lead in African politics. I found myself learning something quite different from what I had come for.

On the Wednesday afternoon of that week, I moderated a dialogue with Rt. Hon. Justice Domitilla Mukantaganzwa, Chief Justice of Rwanda. Before her elevation to the country’s highest judicial office in December 2024, she had served as Executive Secretary of the National Service of Gacaca Courts, the community-based system that adjudicated nearly 2 million cases at the village level in the years following the Rwandan genocide. ‘Justice on the grass’, they called it.

Ordinary Rwandans, elected as lay judges by their neighbours, heard evidence from those who lived within the same hills, delivering judgements that carried weight because the community’s own authority stood behind them.

Justice Mukantaganzwa has written and spoken publicly about what Gacaca taught her about the nature of authority. She describes it as an approach rooted in culture, in dialogue and mediation, as ancient African methods that predate the courtrooms we inherited from colonial legal systems, and in authority that reaches people because it comes from close to where they live.

She has been honest too, about Gacaca’s imperfections. But her fundamental point endures. Where the state cannot easily reach, the community must. And where community reaches, the state must be willing to recognise the legitimacy of what emerges.

I spent much of the week thinking about that idea. I had come to Kigali to think about women’s political leadership. I was leaving with a question that had less to do with women and even less to do with Rwanda specifically. It had to do with Ghana, and with the office I had been appointed to occupy in May 2025, and with a paradox I had been living, but had not yet learned to name.

The paradox is this. Ghana has, over nearly four decades, built one of the most elaborate legal architectures for decentralisation on the African continent. We have District Assemblies. We have a constitutional chapter on local government. We have the National Association of Local Authorities of Ghana (NALAG).

We have the District Assemblies Common Fund (DACF), transferred by constitutional formula. We have a Ministry devoted to decentralisation. We have the Regional Coordinating Councils (RCCs) led by the Regional Ministers. We have a professional Local Government Service (LGS). And every year on the 10th of August, we join the African Union in observing an Africa Day of Decentralisation.

Ghana signed the Charter that established this observance in 2016. Parliament approved ratification in August 2020. The instrument of ratification, however, is yet to be deposited with the African Union. On paper, we are decentralised. In practice, I am beginning to understand, we are something else.

2. Coming Home

I flew home from Kigali to Tema, the city I grew up in.

Every MMDCE’s story is at some level a story about the city or community that made them, and mine is no exception. I am a daughter of Tema. I know the community numbers not from a map but from memory. I know the greenery that once softened every roundabout, the hedges that lined our compounds, the neighbourhood playgrounds where children moved between homes without concern, the clubhouses of VALCO and Ghana Ports and Harbours Authority that anchored the social life of a generation of Tema families, the schools we attended, the streets whose orderly grid was itself part of what it meant to grow up here.

I have seen this city change over decades, in ways that have given me great pride and in ways that have caused me real distress.

Tema was designed to be Ghana’s showcase. Kwame Nkrumah’s industrial capital. A planned city, one of the very few on the African continent, with communities numbered and gridded, drainage engineered, sewerage built, public reservations set aside, roads laid out with intention.

The city was substantially the work of Theodore S. Clerk, Ghana’s first professionally certified architect and the first head of the Tema Development Corporation, working alongside the Greek planner Constantinos Doxiadis, whose firm produced the definitive master plan in the early 1960s for an eventual population of 250,000 within 25 years. The famous Tema numbering system, Community 1 through the higher numbers, comes from that plan.

A separate resettlement town at Tema Manhean, for the fishing community displaced by the new port, was designed by the British modernist architects Maxwell Fry and Jane Drew, whose other work in Ghana includes the university buildings at Kumasi and Achimota. There was a coherence to the original design that a child could feel without being able to articulate. I felt it.

That coherence has degraded over my lifetime. Not all at once. Not through any single failure. Through the slow accumulation of unauthorised structures on land meant for public use. Through drains that stopped being cleared with the regularity the original system required. Through a sanitation architecture that could not keep up with a population that grew far beyond its original design capacity. Through public open spaces quietly taken over for private purposes.

Through greenery that thinned, playgrounds that disappeared, hedges that were replaced with fences and then walls.

Every unauthorised structure, every choked drain, every faded community park is not just a problem to me. It is a chapter in the story of a city I remember being different.

The story of Tema cannot honestly be told without Tema Manhean and Tema New Town.

When the government acquired the 166 square kilometres that became the Tema Acquisition Area in 1952, the original people of Tema were relocated from the site of what is now the harbour to Tema Manhean. The Fry and Drew resettlement plan was, for its time, an unusually participatory piece of design, developed in cooperation with the chiefs.

That care did not carry through the decades that followed. Tema Manhean did not receive the sustained investment that the planned communities of central Tema received. It grew organically, densely, without the same infrastructure, and it has for a long time been what one honest recent chronicler called “the poor cousin of the Tema Township”. That is not the debt we owed to the people whose land became the port that made this a city.

Today, Tema Manhean is losing houses to the sea. At Awudum, at U Compound, at Tooh, buildings hang over collapsing embankments, and entire stretches of coastline have already gone. Two lives were lost in early 2026 when a section of shoreline gave way and a building fell into the sea. The expansion of the Tema Port has been cited by coastal researchers as adding to the pressures already bearing on this stretch of coast. The risk to lives and infrastructure is undeniable, and the community itself has been living the consequences.

And in the last few weeks, Tema New Town was in the national news for a different reason. The volunteer group BuzStop Boys released video footage from the fish drying area showing fish laid out to dry in conditions that were, by any reasonable standard, unsanitary.

The images were shared widely, the Food and Drugs Authority moved quickly, the Assembly and the police joined the intervention; and even the President felt the need to comment publicly on what the country had seen.

Every one of us with any responsibility for Tema had reason to reflect on what those images showed, and on what they revealed about the sustained investment we owe to that community.

When President Mahama nominated me as Mayor in April 2025, and when the Assembly confirmed me on the 9th of May, I felt what I imagine any daughter of a place, who is asked to lead it would feel. I felt an opportunity. I felt the weight of responsibility. And more than that, I felt the genuine thrill of being asked to lead the city I grew up in and love. The moment felt like it carried real possibility. I was going to fix it.

The excitement has not left me. I remain determined, I remain committed, and I remain unpersuaded that our situation is one of despair.

But I have learned something in the 18 months since taking office that no amount of theoretical study of local government and governance could have prepared me for. There is a substantial gap between the will to act and the ability to act.

It is not a lack of will. It is not a lack of vision. It is not a lack of plans. It is not even, in many cases, a lack of resources. The gap is speed. The gap is the ability to activate agency. The gap is between deciding to act and being able to act.

Consider what this looks like in practice.

A resident calls the office to complain about a pothole on their road. The call is legitimate. The complaint is fair. The road is deteriorated. I want it fixed. I have every intention of ensuring that it is fixed.

What the resident does not know, and what I did not fully know before I took office, is what stands between my willingness and the pothole being repaired.

Many of the roads that pass through Tema are not routine local roads. They are national infrastructure carrying the port, the industrial estate, and other assets of national economic significance. Substantive work on these roads, including stretches that read to a resident as ordinary local repair, must be coordinated with the Ministry of Roads and Highways and its Department of Urban Roads, and the funding usually sits above the level at which the Assembly can act alone.

Even where we have the resources, and even where we have the will, the work cannot proceed without national and regional coordination and clearance. That coordination is legitimate. But its pace is not always calibrated to the urgency the resident is feeling. So the pothole waits.

Take another situation. A community decides it wants to fix a shared piece of infrastructure, a drainage channel, or a stretch of neighbourhood road that the composite budget will not reach for another two years.

The Local Governance Act itself, Act 936, empowers the Assembly to impose special rates and levies for exactly this kind of purpose. The community is willing to be levied. The law grants agency.

Yet activating that agency runs through the same architecture designed for larger and higher-risk expenditure. The composite budget process must absorb it. Procurement thresholds still apply. Sector coordination must still be sought.

So the community waits, or gives up, or does the work informally with all the risks that it carries.

Take a third. The Assembly has revenue owed to it by ratepayers whose obligations are, on the face of the law, straightforward.

Some of those ratepayers are ordinary Tema residents and small businesses, and they pay, and they are the ratepayers who keep the Assembly functioning.

Some, however, are large entities whose activities have a profound impact on our infrastructure, our sanitation, our roads, and our environment, and whose willingness to submit to the ordinary jurisdiction of the Assembly is inconsistent.

Some are state-linked. Some are strategic investments. Some have become, in a phrase I have found myself using in private conversations, supposedly too big to touch. Not because the law does not apply to them, but because the political architecture around them makes its application slow, uncertain, and often escalating beyond the level at which the Assembly can readily act on its own.

Tema is often in the news these days, and rarely for reasons that give me pleasure. Sanitation. Drainage. Unauthorised development. The state of the roads. The fish.

Each of these stories is a legitimate story. Each of them reports something the residents of my city are living with. Every news cycle is another reminder that the city needs urgent attention. And I am the Mayor.

And yet what stands between me and the pace this city needs is a pattern of constraints.

The person who runs the finance department reports to the Controller and Accountant-General. Substantive road work depends on ministerial funding and coordination. The composite budget must be approved elsewhere. The procurement threshold requires escalation.

I do not lack authority in principle. I lack operational control in practice.

It is not that I do not want to act. It is that I cannot always activate the speed I know is needed.

The residents deserve to know that this is the honest picture. They do not deserve to be told that everything is fine. They do not deserve to be told that a Mayor with more energy would fix in weeks what has accumulated over decades.

They deserve to be told the truth about the constraints, and they deserve a Mayor who will work to reform those constraints rather than pretend they do not exist.

I sat in Kigali listening to a Chief Justice describe authority that reached people because it came from close to where they lived. I have come home to a city where authority is present in name but attenuated in practice.

Not attenuated by lack of will. Attenuated by the design of the architecture we have inherited.

I have called this paradox, in my own working phrase, ‘Centralised Decentralisation’.

3. A Concept That Ghana Needs

Ghana is not the first country to encounter this paradox, and the scholarly literature on decentralisation has long been aware of it.

In his 1999 study, The Political Economy of Democratic Decentralization, published by the World Bank, James Manor put the point plainly: when the transfer of tasks to lower levels of government is not accompanied by a genuine transfer of authority to democratically accountable local bodies, the result tends in practice to become a more efficient centralisation, not a devolution.

Central actors become better able to project their influence downward, precisely because the local structures they now work through give them a legitimacy they did not previously have.

The tradition of subsidiarity has been the standard response. Familiar in constitutional traditions worldwide and, importantly for our purposes, in African Union documents, including the very Charter we are gathered to observe today, subsidiarity is the principle that decisions should be taken at the lowest level at which they can be taken effectively.

Higher levels retain what they can genuinely do better because of scale, coherence, expertise, or the need to coordinate across boundaries. Lower levels take what they can do better because they know the terrain, they can move quickly, and they are accountable to those who live with the results.

Subsidiarity is not an argument against central government. It is an argument for locating each function at the level where it works best.

Article 6 of the African Charter on the Values and Principles of Decentralisation adopts subsidiarity as a foundational principle. So did the Yaoundé Declaration that preceded it in 2005. So does the European Charter of Local Self-Government.

So do the constitutions of every country I know that has taken devolution seriously. This is not a radical proposition. It is the mainstream position of every serious constitutional democracy that has thought about the design of multi-level government.

The Ghanaian question is what happens when a country has done the formal work of decentralisation but has not achieved the substantive work of subsidiarity.

When the assemblies exist but the levers of action do not fully sit with them. When the community can decide, through its General Assembly, but cannot readily act on its decision. When the local political head is chosen with democratic ritual, but the administrative machinery, the fiscal calendar, the procurement authority, and the presence of parallel powers within the same territory all render that political authority partial.

In what follows, I set out what I mean.

First, where centralisation is not the problem, and can even be the solution. That distinction is a necessary preliminary, because too much of our public conversation about decentralisation proceeds as if centralisation were always suspect, and it is not.

Second, what I have come to call the strings. The invisible levers that, taken together, make the difference between an MMDCE with formal authority and an MMDCE with operational agency. I count six of them.

Third, two dimensions of the paradox that are not visible when one only looks upward from the Assembly to the centre. One is the blocked community, the way a community’s own agency, legally granted by our own statutes, is constrained by the same architecture designed to constrain larger actors.

The other is what I have taken to calling the second enclosure, the presence within our metropolitan boundaries of entities whose activities profoundly shape the city but whose willingness to submit to its ordinary jurisdiction is inconsistent.

Fourth, what other countries have done and what their experience suggests. Kenya has moved further towards genuine devolution than most of us on the continent have. Rwanda has taken a different path, one heavily steered from the centre yet with real local delivery. Uganda offers a cautionary tale.

Fifth, the reform moment we are now in. Ghana is preparing for the most consequential set of local government reforms in a generation. A Constitutional Review Committee has reported. A government position paper has responded. A referendum is projected.

I set out these positions honestly and non-partisanly, because I take them seriously, and because none of them, on its own, will resolve the paradox I describe. The strings run deeper than the question of how the political head is chosen.

Finally, what genuine subsidiarity would look like in the Ghanaian context.

This is a long argument. I hope you will read it patiently. It is what I have been living for 18 months, and I would like to share it before another Africa Day of Decentralisation passes with the paradox still unnamed.

4. Where Centralisation Belongs

Not all centralisation is a problem. Ghana is a unitary state, and there are functions that only a central actor can do well.

National curriculum. Clinical protocols. Environmental standards. The national spatial framework. Macroeconomic policy. Foreign policy.

No serious observer suggests these should sit with 261 assemblies deciding separately, and I do not either. The framework at the centre is what makes a country coherent.

The framework I want to draw on comes from subsidiarity theory, grounded in my Tema experience. It has two dimensions:

Dimension 1: What kind of function is it?

Framework functions (should be central): National standards, curriculum, clinical protocols, regulatory frameworks, quality assurance, cross-jurisdictional coordination, national spatial plans, environmental protection standards.

Delivery functions (should be local): Decisions about implementation, service delivery, day-to-day operations, adaptation to local context, responsiveness to local problems, local zoning within the national framework, and spending decisions of Internally Generated Funds (IGF) and, to some extent, utilisation decisions of the District Assemblies Common Fund (DACF) allocations.

Dimension 2: What kind of control is it?

Ex-ante controls: You need permission before you can act. Slow but predictable. Designed to prevent misconduct before it happens.

Ex-post accountability: You can act, but you will be audited and held accountable afterwards. Fast, but requires strong audit and oversight institutions.

Ex-ante controls require permission before action. Ex-post accountability lets you act and then audits, questions and sanctions after the fact.

Ghana operates both, and this is not accidental. Both were designed to guard against corruption and the abuse of public resources, which is a legitimate purpose that any serious reform must preserve.

Alongside the ex-ante permissions we clear before we act, we now answer to a substantial ex-post infrastructure that includes the Auditor-General, the Public Accounts Committee of Parliament, the Public Interest and Accountability Committee, the Right to Information Act, and the Office of the Special Prosecutor.

These institutions are real, they are stronger than they once were, and they are a welcome development for our democracy.

Assemblies do have real authority under the Local Governance Act 2016 (Act 936), and we use it every day. The subtler problem is one of balance.

A country trying to move fast cannot afford a control architecture calibrated for a country moving slowly. The anti-corruption logic behind ex-ante control was defensible when it was designed.

But if our ex-post institutions are now capable of catching abuse after it happens, we can afford to loosen the ex-ante grip on willing and capable actors without dismantling the accountability we need.

The willing MMDCE clears the ex-ante layer before acting, and answers to the ex-post layer after acting, and each takes its own time. The delivery and the resident waiting for it sit between them.

This is the principle against which I want us to measure Ghana’s current arrangements. Not against a naive assumption that all decentralisation is good and all centralisation is bad. Against the honest test of whether each function sits at the right level and whether each control operates at the right point.

Measured against that test, we have work to do.

5. The Strings We Do Not See

Six inherited strings shape the pace of what an assembly can actually do. They operate in similar ways in every one of Ghana’s 261 assemblies, and they are cross-partisan, structural, and long-standing.

Any of them, taken alone, has a defensible rationale. Their cumulative weight is the problem.

The First String: The Administrative Head Is Not the Assembly’s Employee

At the top of every MMDA sits a Coordinating Director, the administrative head of the entire assembly machinery. In Tema, this is the Metropolitan Coordinating Director.

This is the officer who runs the day-to-day administration, serves as Secretary to the General Assembly, is the principal spending officer for the composite budget, and supervises the heads of every department in the Assembly.

The Coordinating Director is not an employee of the Assembly. Under the Local Government Service Act 2003 (Act 656) and the Local Governance Act 2016 (Act 936), the Coordinating Director is an officer of the Local Government Service, appointed, posted, promoted, disciplined, and transferred by the Service and its Council.

The elected MMDCE is described in the statute as the person to whom the Coordinating Director provides secretariat services and special advice, but the MMDCE has no formal appointing, disciplining, or transferring authority over the officer who runs the administration they lead.

The Assembly has a District Appointments Committee under Act 936, which handles appointments and disciplinary matters for other staff. That committee is chaired by the Coordinating Director. Its recommendations require the prior approval of the Head of the Local Government Service.

The advisory character is real. The authority is elsewhere.

Professional public services around the world separate political leadership from career administration precisely to insulate the administration from political interference. That is a legitimate objective, and it should be preserved.

What is at issue is the difference between insulation and disconnection. The current arrangement gives the political head no meaningful say over the officer whose performance most directly determines whether the political head’s vision is implemented.

The Coordinating Directors I have worked with have generally been professional, dedicated public servants doing difficult jobs with limited resources. This is not a criticism of the people. It is an observation about the structure.

The Second String: The Finance Department Answers to the Controller and Accountant-General

If the first string constrains the assembly’s administrative leadership, the second string constrains its financial operations, and does so more comprehensively than most residents would imagine.

Under the Public Financial Management Act, 2016 (Act 921), the Controller and Accountant-General is the Chief Accounting Officer of the Government of Ghana and is statutorily responsible for providing accounting officers to covered entities, which include MMDAs.

The Metropolitan Finance Officer, the officer who runs the finance department at every assembly, is an employee of the Controller and Accountant-General’s Department.

That officer sits physically in the assembly’s offices. Uses the assembly’s furniture, air conditioning, and equipment. Interacts daily with the political and administrative leadership of the assembly.

But their employer, their disciplinary line, their promotion pathway, and their professional loyalty all sit with the Controller.

If a finance officer performs poorly, the Assembly cannot place a query on its own records. The Assembly does not hold their personnel file.

Any disciplinary process must be routed to the Controller and Accountant-General’s regional office and, if serious, to the Controller. It is a slow process by design, and the officer is aware, at every step, that the assembly’s leadership is not the source of consequence in their career.

This is the string I find most consequential, because everything else the assembly does runs through the finance function. Every project payment, every rate collection, and every audit response passes through officers whose primary employer is the Controller and Accountant-General Department.

The Third String: The Roads Are Not Ours to Repair

The Department of Urban Roads, established in 1988 under the Ministry of Roads and Highways, is responsible for urban road networks across Ghana’s cities.

Unlike a District Assembly, where roads sit as a unit inside a merged Works Department, a Metropolitan Assembly like Tema has, under Legislative Instrument 1961 of 2009, a dedicated Urban Roads Department with its own professional head.

That department reports administratively to me through the Coordinating Director, and my experience so far in Tema is that the staff have been responsive and professional.

The constraint is not with the people. It is with what sits above them.

The engineering standards they must follow are set nationally through the Ghana Standard Specifications for Roads and Bridges. Their technical direction is provided by the regional office of the Department of Urban Roads. And the funding for anything substantial comes from the Ghana Road Fund and the Ministry of Roads and Highways, not from the Assembly’s own resources.

Many of the roads a resident would call “Tema roads” are, in national terms, strategic economic infrastructure in which ministries and agencies beyond the Assembly have legitimate interests.

That produces a real coordination requirement, and I want to be candid about it. When port traffic patterns shift or industrial demand changes, the roads must respond to more than the residents living beside them.

That requires the Ministry of Roads and Highways, the Department of Urban Roads, the Ghana Ports and Harbours Authority, and others to have meaningful input into what happens on those roads.

This is not, in itself, a design flaw. It is a legitimate consequence of Tema’s economic character.

The consequence, however, is that even where the Assembly identifies a road that needs work, the cost is often beyond what our composite budget can absorb, and the project must be routed through the ministry for funding and technical review.

Substantial road interventions typically sit above the fiscal threshold at which an Assembly can act on its own authority.

But the pace of delivery for anything substantial is set at the level of regional and national coordination, not at the level of local urgency.

Potholes, shoulder repairs, and drainage clearances that a resident reasonably expects to see dispatched within a season instead move at the pace of the fund release and the Department of Urban Roads timetable.

The Mayor cannot commit to a residents’ association that a road will be fixed by a given date, because the Mayor does not set the priorities at the regional and national levels.

The Fourth String: Some Sector Departments Belong Somewhere Else

The pattern I have described for finance and for urban roads is not unusual in the assembly structure. It is closer to being the rule.

Across most of the departments an assembly is meant to coordinate, the substantive employer is a national agency or a regional structure, and the reporting line runs elsewhere.

The Metropolitan Education Directorate is staffed by Ghana Education Service employees. The Metropolitan Health Directorate is staffed by Ghana Health Service employees.

Physical Planning staff sit within a framework directed by the Land Use and Spatial Planning Authority under the Land Use and Spatial Planning Act 2016 (Act 925). Statistical Service staff report to the Ghana Statistical Service.

Agriculture staff, though formally devolved under earlier reforms, still operate within a strong Ministry of Food and Agriculture framework.

The Business Advisory Centre, which supports the Assembly’s Local Economic Development function, has its activities and reporting authority directly accountable to the Ghana Enterprises Agency under the Ministry of Trade, Agribusiness and Industry.

Under Legislative Instrument 1961 of 2009, the departments of an MMDA were formally absorbed into the assembly structure. The absorption was administrative. The employment was not.

The staff continued, in most cases, to be employees of their national parent agencies, and the reporting lines have not fundamentally changed.

The Assembly is expected to coordinate the composite work of these departments, to submit a composite budget that integrates their plans, and to be held publicly accountable for the outcomes.

It does not employ them.

There has been movement on this. The Ministry of Education, at the direction of President Mahama, has initiated work on a bill to devolve elements of teacher recruitment, posting, and sanctioning to the assembly level.

The Inter-Ministerial Coordinating Committee on Decentralisation is working with the ministry on the drafting.

This is significant, and I welcome it. Whether it delivers a genuine transfer of the employer function, or a re-labelling of reporting lines that leaves the substantive employer at the centre, remains to be seen.

If it succeeds, it becomes the template for health, for agriculture, and beyond. The stakes are high.

Beyond the sector departments themselves, the Assembly’s engagement with the ministries flows through a further layer.

Under Section 188 of Act 936, each region has a Regional Coordinating Council, chaired by the Regional Minister, whose functions include monitoring, coordinating, and evaluating the performance of the District Assemblies in its region, approving assembly by-laws subject to their consistency with national legislation, and reviewing and coordinating public services generally in the region.

The Regional Minister is a central government political appointee.

It performs coordination that a decentralised system does need. But it is another layer through which much of the Assembly’s engagement with the wider central architecture must flow, and every layer takes its own time.

The Fifth String: The Procurement Threshold and the Fiscal Calendar

The fifth string binds not just the Assembly’s independent action but its capacity to act at speed on its own priorities, and it operates in two related ways.

The first is procurement.

The Public Procurement Act 2003 (Act 663), as amended, and the Public Procurement Regulations 2025 (L.I. 2516) organise procurement in tiers.

Assemblies handle procurement below defined ceilings on their own authority. The most recent regulations have raised those ceilings in real terms, which is a genuine improvement.

What remains above them is a tiered escalation to review boards for higher-value investments into the community.

This is the point at which the architecture becomes most consequential for the pace of local delivery, because it is precisely the transformational projects, the ones a community most needs to see completed within an electoral cycle, that most often sit above the ceiling.

A single major investment, whether it is a treatment plant, a market upgrade, a truck park, or a coastal defence, enters the escalation process.

What began as an Assembly commitment to a community becomes a national approval process.

The escalation is not just slow. It changes the character of the decision, because the timing and the review no longer answer to the community that will live with the result.

The second is the fiscal calendar.

The District Assemblies Common Fund is a constitutional transfer under Article 252 of the 1992 Constitution. It is the single largest source of development finance for most assemblies.

And it has, for many years, arrived late and depleted, sometimes so late that the Assembly cannot commit to the projects it planned within the year of the plan.

The DACF Administrator himself has called for a statutory release timetable to be locked into law.

The government’s own budget commitments have improved the direct transfer share, moving it towards 80% direct to assemblies, and this is a welcome improvement, I acknowledge.

But it does not resolve the fundamental point. The Assembly plans on money whose arrival it cannot predict, and residents live with the consequences of the delay.

The composite budget, once prepared, is reviewed and approved through a national process. By-laws that the Assembly enacts to govern its own jurisdiction require ministerial oversight under Act 936.

The fiscal string is the one that most directly ties the political head’s hands.

An MMDCE cannot commit to a project whose funding will not arrive on time. And the Assembly cannot commit to the transformational investments a community actually needs, because the moment the ambition scales, the decision escalates.

The Sixth String: The Plan Is Composite in Name and Vertical in Fact

The sixth string ties all the others together.

Under the National Development Planning (System) Act 1994 (Act 480) and the frameworks issued by the National Development Planning Commission, each assembly is designated as the planning authority for its area.

The Assembly is required to prepare a Medium-Term Development Plan and an Annual Action Plan that integrate the plans of all its departments and align with the national development framework.

In practice, the assembly’s plan integrates the vertical plans of the decentralised departments and agencies, but in some cases does not direct them.

The Ministry of Education runs its programme of school construction and teacher deployment according to its own criteria. The Ministry of Health runs its programme of facility upgrading and health worker deployment according to its criteria.

The Ministry of Food and Agriculture and the Ministry of Roads and Highways each run their own vertical streams.

In most cases, these ministries carry out projects and activities within the Assembly’s jurisdiction without recourse to the Assembly’s Annual Action Plan, and even where their departmental representatives sit on the Assembly and contribute to the composite planning process, the ministries often operate off separate national plans that the Assembly is not fully sighted on.

When residents ask why a school was not built where they asked, or why a road was not repaired on the timeline they needed, the honest answer is often that the decision was not the Assembly’s to make.

The Six Strings, Held Together

Look at these six strings together, and you see the shape of the paradox.

The administrative head is not the Assembly’s employee. The finance department whose command line runs to the Controller and Accountant-General. The urban roads whose delivery pace is set by the Ministry and the Department of Urban Roads, not by the Assembly.

The sector departments whose employees answer elsewhere. The procurement thresholds and fiscal calendar that gate independent action. The plan that is composite in name and vertical in fact.

None of these strings, on its own, renders an assembly powerless. Any of them, taken alone, could be defended on legitimate grounds.

Together, they produce the pattern I have been describing.

An MMDCE with formal authority who commands, in operational terms, remarkably less than the office would suggest.

A political leadership that is answerable for outcomes it cannot fully control.

A community that is asked to trust institutions whose lines of accountability are not visible to it and are frequently not accountable to it.

This is what I mean by Centralised Decentralisation.

It is not the absence of a local government system. It is the presence of a local government system whose substance has been drawn out of it, string by string, over decades of design choices that each had defensible reasons but whose cumulative effect is a paradox.

6. The Blocked Community

There are two further dimensions of Centralised Decentralisation that only become visible when you look sideways rather than upward.

The strings I have already described run vertically, from the Assembly toward the centre.

This one runs inward, from the Assembly toward its own residents.

The next runs horizontally, from the Assembly toward parallel actors operating within the same city.

The Local Governance Act 2016 (Act 936) makes the Assembly the sole rating authority for its district. Section 146 permits the Assembly to levy general or special rates, and where the Assembly moves to impose a special rate, the statute requires it to consult with district-level stakeholders.

This is one of the most underused subsidiarity mechanisms in our whole local government architecture.

It contemplates exactly what a community should be able to do.

A community identifies a shared priority that the composite budget will not reach in time. The Assembly, in consultation with that community’s stakeholders, levies a special rate to fund it.

The community pays. The project is delivered. Local democratic agency is exercised at the level closest to the residents who will live with the result.

The blockage comes after the levy is agreed.

Once the Assembly has done the consultation and imposed the special rate, the delivery of the project runs through the same operational architecture that processes every other assembly activity.

The composite budget process must absorb it. The procurement tiers still apply. The relevant sector coordination must still be sought.

The finance function and the administrative machinery process it at the pace of the wider system, not at the pace the community consented to when it agreed to be levied.

At that rate, it would take years to complete what a community expected to see within months.

Subsidiarity holds that decisions should be taken at the lowest level at which they can be taken effectively.

A community consenting to a rate levied for its own priority is subsidiarity at its purest.

When our operational architecture cannot match that decision with delivery, we are not merely slowing a project.

The special rates provision is one of the few places in our statutes where the community’s own priority is treated as a legitimate source of local action.

If we cannot deliver on it at the pace the community expects, we have blunted a tool that decentralisation is meant to sharpen.

7. The Second Enclosure

The third geometry runs horizontally, from the Assembly toward other state and quasi-state entities that operate within the city’s boundaries but outside the city’s authority, and the central government’s tolerance of this weakens the assembly further.

This is perhaps the least discussed feature of Centralised Decentralisation, and it may be one of the most consequential.

The Tema Metropolis hosts several major industrial and commercial entities whose activities generate a significant impact on our infrastructure, but whose willingness to submit to routine local jurisdiction, from access for inspection to timely rate settlement, is inconsistent.

They act, in effect, as if they are answerable elsewhere. And in a real sense, they are, because the central government treats them as strategic assets rather than as ratepayers subject to the ordinary authority of the city they operate in.

Some are state-owned enterprises. Some are major private investments. Some are former statutory bodies now converted to limited liability commercial vehicles that continue to conduct themselves as if the earlier statutory authority remains vested in them.

The most consequential example for Tema is the Tema Development Corporation.

TDC was created as a statutory body to hold, develop, and manage the Tema Acquisition Area on behalf of the state.

Parliament removed its regulatory planning authority when it passed the Local Government Act 1993 (Act 462), and reinforced that removal through the Local Governance Act 2016 (Act 936) and the Land Use and Spatial Planning Act 2016 (Act 925), both of which vest planning authority exclusively in District Assemblies.

In 2017, TDC was converted from a statutory body into a limited liability company, cementing what its role should now be.

That role is commercial, focused on managing and monetising the state’s land holdings, similar in principle to how other state-owned commercial entities like ECG operate within their spheres.

Yet TDC continues to conduct itself in significant respects as if the earlier statutory authority remains vested in it.

Development permits, enforcement notices, public land allocation decisions, and land use directions are still issued by TDC across the Acquisition Area, even though the statutory planning authority for that territory is the Tema Metropolitan Assembly and its neighbouring assemblies.

Two authorities, in effect, occupy the same jurisdiction.

One by statute. The other by inheritance, habit, and the central government’s failure to enforce the settled statutory position.

The consequences of the wider pattern are concrete.

A city cannot plan coherently when significant land areas within its boundaries are effectively outside its planning authority.

A city cannot fund itself when significant ratepayers within its boundaries treat their obligations as negotiable.

A city cannot enforce environmental or development standards uniformly when some actors are subject to the ordinary process, and others are, in effect, exempt.

The residents see this. They know when the same rule is applied to some but not to others.

And when they lose confidence in the evenness of local authority, they lose confidence in local authority itself.

Centralised Decentralisation therefore operates along three axes.

The vertical axis, in which the central government retains operational control of the Assembly’s staffing, finance, sector functions, procurement, and planning.

The community-inward axis, in which the community’s own agency, legally granted, is throttled by the same architecture that gates every other assembly action.

And the horizontal axis, in which parallel actors within the city’s boundaries are effectively insulated from local jurisdiction by the central government’s tolerance of their exception.

It is not just that the Mayor cannot fully command the administration.

It is that the community cannot fully activate its own agency.

And it is that entities operating in the city are not fully subject to the city’s authority.

The result is a jurisdictional authority that looks complete on paper but is fractured in practice.

8. What Others Have Done

Ghana is not the first country to encounter this paradox, and the responses of other African democracies are instructive.

Three cases in particular illustrate the range of choices available and the consequences of each.

Kenya represents the most substantive attempt at devolution on the continent.

The 2010 Constitution created 47 counties, each with an elected Governor and a County Assembly. Article 203(2) guarantees the counties not less than 15% of nationally raised revenue, transferred by formula rather than by discretion.

Article 204 establishes an Equalisation Fund for the 14 most marginalised counties.

And, decisively for the argument I am making, Article 235 requires each county to maintain its own public service, run by a County Public Service Board that hires, disciplines, and removes its staff.

The Governor is elected. The County Public Service Board sits with the county.

The devolution is fiscal, administrative, and political.

Whatever difficulties the Kenyan system continues to face, and there are many, the structure has produced something quite different from what Ghana currently has.

If a country devolves the political office without devolving the employer function and the fiscal calendar, the political office holder inherits an administration they cannot fully command and a budget they cannot fully predict.

If it devolves all three, the local authority becomes something recognisably close to what decentralisation is meant to be.

Rwanda offers a different lesson.

The Rwandan model is heavily centrally steered but locally effective. The country is organised into 30 districts, and each district signs an annual Imihigo performance contract with the President, publicly, with specific targets on service delivery, revenue, education, health, and economic development.

District mayors are indirectly elected by their councils but are tightly bound by the Imihigo they have signed.

The centre sets the framework and the targets. The local level delivers with real accountability for whether the targets are met.

This is not a model of autonomous local government. It is a model of disciplined implementation within a strong central steer.

It works in Rwanda for reasons that include the country’s specific history, its size, and the exceptional discipline of its central administration.

It demonstrates that local effectiveness does not require full autonomy. It requires clear mandates, real accountability, and a central actor that treats local performance as a strategic priority.

Ghana’s problem is not that the centre is too strong. Ghana’s problem is that framework and delivery are tangled at every level, and the pace suffers.

Uganda is the cautionary tale.

Uganda began its decentralisation reforms in the early 1990s with genuine ambition. Local councils were established. Fiscal transfers were designed. The system was seen, for a time, as one of Africa’s more promising devolutionary experiments.

Then it was progressively rolled back.

A 2005 constitutional amendment returned the appointment of Chief Administrative Officers, the equivalent of our Coordinating Directors, to the central Public Service Commission.

Payroll administration was recentralised. Drug procurement for local health facilities was recentralised.

The graduated tax, which had been the main source of local revenue, was abolished before the 2006 elections.

By the middle of the following decade, Ugandan local governments were dependent on the centre for approximately 95% of their funding.

The formal architecture of decentralisation remained. The substance had been drawn out of it.

What Parliament gives, Parliament can take back, and central actors will find reasons to take it back if the entrenchment is not secure.

Three cases, three lessons.

Kenya shows what full devolution of the political, administrative, and fiscal functions looks like when it is constitutionally entrenched and formula-driven.

Rwanda shows that disciplined local delivery is possible even under strong central direction, provided the framework and delivery lines are clean, and the accountability for performance is real.

Uganda shows what happens when devolution is not entrenched and can be rolled back one recentralisation at a time.

Ghana’s current position sits somewhere between the three, drawing elements from each.

We have the formal architecture of Kenyan-style devolution without the fiscal or administrative substance.

We have some of the framework and delivery ambitions of Rwanda without the discipline or the performance accountability.

And we have Uganda’s vulnerability, in the sense that our devolution has never been fully entrenched and remains reversible in ways that Kenya’s is not.

The question for the reforms now on the horizon is which of these paths Ghana chooses.

9. The Reform Moment

Over the last decade, three positions have defined the debate about how MMDCEs should be chosen.

The first is the position advanced by the NPP administration between 2016 and 2019. It proposed amending Article 243(1) of the Constitution to elect MMDCEs and, separately, amending Article 55(3) to permit political parties to participate in local government elections.

The referendum scheduled for 17 December 2019 was cancelled, and both bills were withdrawn.

The second is the position of the Constitutional Review Committee chaired by Prof. H. Kwasi Prempeh, which submitted its report to President Mahama on 22 December 2025.

The CRC proposed the direct election of MMDCEs on a non-partisan basis, with no presidential nomination role in the candidate selection.

It proposed a phased rollout across the 261 districts, with an Independent Devolution Commission determining the sequence and the criteria for districts to qualify for the vote.

It recommended stricter qualifications for MMDCE candidates given the executive character of the role, and it proposed raising the constitutional minimum for the District Assemblies Common Fund to 10% of national revenue.

The third is the Government Position Paper announced by the Attorney-General, Dr Dominic Ayine, on 30 July 2026 as the current NDC administration’s response to the CRC report.

The government endorsed elected MMDCEs and adopted a mechanism under which the President nominates 5 candidates for each position, including 2 women, of whom 3, including at least 1 woman, are shortlisted by a vetting committee comprising representatives from the Ministry of Local Government, the Public Service Commission, and the Local Government Service.

The 3 shortlisted candidates then contest the election.

The government declined the CRC’s phased approach in favour of simultaneous elections nationwide.

The bills implementing these amendments are targeted for October 2026, and the referendum on the entrenched provisions is planned to run alongside the 2027 District Assembly elections.

The government has also proposed raising the DACF minimum to 7.5% of national revenue, below the CRC’s 10% recommendation.

The CRC position appears to offer the most complete devolutionary framework of the three.

Even so, the reform intended to devolve the political head retains, in its own design, the shape of the paradox this argument has been describing.

None of the three positions, on its own, will resolve the paradox.

The strings run deeper than the question of how the political head is chosen.

An MMDCE elected under any of these three models would inherit an administration whose head is not the Assembly’s employee, a finance department whose command line runs to the Controller and Accountant-General, roads whose delivery depends on coordination outside the Assembly structures, sector departments whose employees answer elsewhere, procurement thresholds and a fiscal calendar that gate independent action, and a composite plan the Assembly does not truly command.

The election of the political head is a necessary reform. It is not, on its own, a sufficient one.

The next question, therefore, is not how MMDCEs will be chosen.

It is what an elected MMDCE will be able to do once in office, and how the strings that currently bind the appointed office will be loosened, or not, when the office becomes an elected one.

That is the question I want to close on.

10. What Comes Next

Genuine subsidiarity in the Ghanaian context does not require abandoning all centralised control or transferring every function to the Assembly.

It requires a careful relocation of specific functions and a careful rebalancing of specific controls, so that framework work stays at the centre and delivery work moves close to the people who will live with the results.

For an MMDCE to have real operational agency, several things would need to shift together.

The Coordinating Director does respond to the political head in practice, but that responsiveness is largely a matter of courtesy and convention rather than binding structure.

Real accountability sits with the Local Government Service.

A more robust framework would give the Assembly a substantive appointing and disciplinary role over the Coordinating Director, alongside the Service’s role in professional standards, so that political accountability and administrative command are aligned rather than merely coexistent.

The finance department would need to be brought into a similar arrangement.

The Metropolitan Finance Officer and some staff of the finance department answer to the Controller and Accountant-General for professional standards, and would need to answer to the Assembly for the responsiveness and pace of the Assembly’s financial operations, with real consequences for both lines.

The sector functions would need to move the employer status, not just the reporting line.

The education devolution bill under preparation is the test case. If it succeeds in transferring the employer function to the Assembly level, it becomes the template for health, agriculture, and roads.

If it re-labels reporting lines while leaving the employer at the centre, it will not change what an MMDCE can actually command.

The question of procurement thresholds deserves fresher thinking.

If we have built a strong ex-post accountability architecture, and we have, the case for tiered ex-ante escalation of procurement above certain values weakens considerably.

A rethinking of the threshold structure, in light of what our ex-post institutions can now do, would begin the recalibration this argument has been building toward.

The District Assemblies Common Fund would need to arrive on a statutory calendar, as the Administrator has himself proposed, so that the Assembly can plan on money it can predict.

The special rates provision under Act 936 is, at present, largely unused.

Even in situations where a community shows agency and the Assembly decides to activate a special rate for a community-declared priority, the delivery of the funded project moves through the same slow architecture that processes every other Assembly activity.

The instrument needs to be brought closer to the people it was designed to serve.

That means a lighter operational track for community-consented, community-levied work, so that the pace of delivery matches the pace at which residents have agreed to be levied.

The second enclosure would need to be addressed by aligning the actors within a metropolitan boundary to the statutory authority the law has vested in the Assembly, so that no ratepayer, no parallel body, and no strategic investment sits outside the ordinary jurisdiction of the city that stewards them.

None of these shifts abolishes the framework role of the centre.

Each of them locates a specific delivery function at the level where subsidiarity says it belongs.

And each is achievable within the reform window Ghana is now approaching, provided we make the choice to attempt it.

11. My Closing Reflection

I began in Kigali, in the company of a Chief Justice whose earlier work led one of Rwanda’s most difficult efforts of reconciliation at the village level, through elected lay judges who sat with their neighbours on the hillsides and rendered judgements that carried the community’s own authority behind them.

What has stayed with me from that week is not the specifics of Gacaca. It is the underlying principle.

Authority reaches people because it comes from close to where they live.

I have come home to a city I love, where I have spent 18 months learning how much harder it is to act than to decide.

The determination has not diminished. What has grown is my understanding of the constraints.

The invisible levers I could not see from outside the office are now the levers I work with, and around, every day.

None of this is a complaint.

I took the office knowing that reform is slow and that transformation is harder than proclamation.

What I have set out here is not a claim that Tema’s problems are the fault of the architecture rather than the responsibility of the Mayor.

They are my responsibility, and I own them.

But the residents deserve to know the truth about the constraints, and the country deserves an honest conversation about what those constraints are, from those of us who are inside them.

Ghana is entering the most consequential period for local government reform in a generation.

The elected MMDCE reform, however designed, will only realise its promise if the strings that currently bind the office are loosened along with the change to how the office is filled.

The framework functions can remain at the centre. The delivery functions must move close.

The ex-ante controls that guard against abuse can remain, but must be recalibrated as the ex-post accountability institutions have strengthened.

The community’s own agency, granted by our own statutes, must be met with an operational architecture that can deliver at the pace the community consented to.

The actors within our cities must be brought under the ordinary authority of the cities that host them.

The 10th of August was set aside by the African Union as a day on which the continent observes the values and principles of decentralisation.

I offer these reflections not as a complete answer, but as an invitation to hold the conversation openly.

I write as a Mayor. I would like Ghanaians to hear this as the beginning of a conversation about what our local government system was meant to be, what it has become, and what, together, we might yet make of it.

. . .

Ebi Bright is the Metropolitan Chief Executive of the Tema Metropolitan Assembly. She writes in her personal capacity as a practitioner and student of local government.

Note on Sources

All statutes are cited to their published short titles as gazetted by the Government of Ghana.

• The scholarly reference to Manor (1999) is to James Manor, The Political Economy of Democratic Decentralization (Washington DC: The World Bank, 1999).

• The Constitutional Review Committee position referenced is drawn from the report submitted to President Mahama on 22 December 2025, chaired by Prof. H. Kwasi Prempeh.

• The Government Position Paper is the response presented by the Attorney-General, Dr Dominic Ayine, on 30 July 2026.

The comparative material in Section 8 is based on the following:

• The continental framing follows Stephen N. Ndegwa, Decentralization in Africa: A Stocktaking Survey (Africa Region Working Paper Series No. 40; Washington DC: The World Bank, 2002).

• The Kenyan discussion draws on the Constitution of Kenya, 2010, Articles 203, 204 and 235 in particular, together with Nic Cheeseman, Gabrielle Lynch and Justin Willis, “Decentralisation in Kenya: the Governance of Governors” (Journal of Modern African Studies, 54:1, 2016).

• The Rwandan account of the Imihigo performance contracts is drawn from Rwanda’s Ministry of Local Government (MINALOC), with scholarly corroboration from Matthew Sabbi and Jean-Baptiste Ndikubwimana, “Innovating Imihigo” (Canadian Journal of African Studies, 58:2, 2024).

• The Ugandan account follows Frederick Golooba-Mutebi, “Politics and Local Government in Uganda,” in F. Saito (ed.), Foundations for Local Governance (Heidelberg: Physica-Verlag, 2008), the figure of over ninety-five per cent central-transfer dependence being that of Uganda’s Local Government Finance Commission (2012).I’ve kept the substance intact and focused on spacing, section hierarchy, paragraph breaks, and readability rather than rewriting the argument.