by Martin van Staden
Introduction
The ultimate purpose of government is the protection of legal subjects’ freedom of action and private property. This is the original social contract, and to the extent that the social contract has since ostensibly been enlarged, this ultimate purpose remains its raison d’être. When considering key constitutional questions like the amalgamation of states or the breaking up of a state, the answer lies with the avenue that best achieves this purpose.
In other words, would uniting the states of Southern Africa into a federal union be better for legal subjects’ freedom and property rights, or worse? If it is worse, then the question need not be further considered.
Federalism versus Sovereignty in International Relations
Centralised and decentralised political orders
Whether a country adopts a centralised or a decentralised constitutional form could entail significant consequences for freedom of action and property. Under a unitary arrangement, it is necessary for a new political elite to only take control of the central government. Once this has been achieved, the policies and laws adopted at the centre would be uniformly applicable across the country. Similarly, having political authority centralised in this fashion amounts to a concomitant centralisation of risk. If a natural disaster were to strike the capital city and kill or decapacitate the leaders of the unitary authority, the remainder of the country could be left in lawless chaos. Under a unitary dispensation there is a single major point of failure.
On the other hand, a decentralised political model – whether this is unitary devolution or constitutionally entrenched federalism – spreads risk out, and diffuses power. Unitary devolution would mean the unitary authority at the centre seeks willingly to decentralise power and does so through discretionary grants of power to new centres of authority across the country. Federation, on the other hand, means various centres of power are created in a legally supreme constitution, meaning the central government cannot, in theory, use its discretion to revoke political authority from other spheres of government.
If adopted in a somewhat sincere manner, federalism ensures that any political elite seeking total control over a society would need to seize multiple centres of power. Should they fail to do so, not only would total control elude them, but the competing political elite(s) that control the remaining centres of power act as a check and balance on the abuses and impulses of the elite at the centre.
Sovereignty and decentralisation
Liberal political theory has always prized decentralisation, however it is a relatively recent phenomenon for liberals to insist somewhat uncompromisingly upon the sovereignty of separate territorial and nation-states.
The notion of sovereignty was borne out of feelings of national consciousness by people-groups united around ethnicity, religion, and/or language. These nations, it was thought, in order to be collectively “free,” must govern themselves, with only voluntary interaction across national borders. Liberalism, far older than nationalism, did not originally conceive of political society in this fashion. The liberal political theory has always been a universal one. Governments are established, anywhere, in order to safeguard the pre-existing natural rights of legal subjects. In return for this guarantee, legal subjects sacrifice only their executive power – the power of self-help, or to enforce their own liberty against others – to government, so that ordered freedom under law, or civil liberty, replaces natural liberty. In other words, to liberals, what matters is whether the government is engaged in the protection or infringement of civil liberty. Various constitutional mechanisms were developed to make the former rather than the latter more likely, among them the trias politica, checks and balances, federalism, and an independent civil society as its own centre of power. Whether a nation is “governing itself” is at best a secondary priority, because having separate territorial or nation-states with their own political sovereignty does not truly act as an additional check and balance upon the power or scope of any individual government.
Checks and balances only operate effectively within a single polity. In other words, whereas the central government and the provincial governments of South Africa, existing side by side within a single polity, do act as checks and balances upon one another, the governments of Zambia or Mozambique do not act as a check and balance upon the South African government or any of its provinces. There has been much rhetoric around how international organisations play a role in “global constitutionalism” however this ultimately fails due to the practically voluntary nature of international law. For example, the moment it became politically inconvenient for South Africa to be a member of the Rome Statute of the International Criminal Court, that government announced its intention to withdraw from the ICC. While international organisations and agreements might play a meaningful role in applying pressure upon sovereign governments, their role in in fact checking power is negligible.
What then could be a coherent liberal contribution to the theory of international relations that does not simply ride the coattails of nationalistic international theory, but rather brings to the fore the liberal concern for a limited government dedicated to the protection of individual freedom of action and property?
We have established that separate “peaks” of sovereignty compartmentalised in separate polities does not have the effect of checking power, whereas entrenching constitutional checks on power within a single polity in fact does. It would seem, then, that the rhetoric of “centralisation” and “decentralisation” applies within polities, not between them. Having 100 separate polities arranged next to one another is not a “decentralisation of power” because within each polity, and as far as the legal subjects there are concerned, power is centralised in whatever their sovereign institution(s) are. However, were these 100 separate polities to become 100 centres of power within a single polity, one can honestly claim that there is a decentralisation of power, because now these 100 centres of power no longer represent power compartmentalised and intensely focused on the legal subjects within their jurisdiction, but power checking and balancing the power of every other former polity.
Uniting states into a single polity therefore brings more people under the protective net of a constitutional system of checks and balances. Such a move could however also be oppressive unless the constitutional system itself emphasises decentralisation and limited government.
Federalism is the marriage of decentralisation and limited government, because a federal relationship between a central and subcentral government requires constitutional definition, and constitutional definition entails political and judicial contestation. Whereas mere amalgamation or unification is therefore quite a risk, federalisation – while not riskless – is the safer route, and could arguably be better protective of freedom of action and property of legal subjects than the present system of separate peaks of sovereignty.
To favour federalism over sovereignty is therefore not to favour centralisation, but in fact to favour decentralisation. Federal unification of polities arguably advances decentralisation more than sovereign independence between polities does.
Southern African Federalism
What is today the Republic of South Africa has a rich federalist tradition stretching back at least to the 1880s. The early federalists were primarily concerned with questions of economy, which would have benefited the colonial authorities by reducing the costs of government, and benefited the settler populations by eliminating what were in effect pointless political boundaries to trade. As time went on, the stakes of federalism increased. By the end of the first decade of the 1900s, when the four colonies that preceded South Africa unified, the question of federalism started to become entangled with the new Union’s racial politics. First, federalism featured as a topic of contention between the British and Afrikaner “races” as late as the 1960s. Second, from at least the 1950s, federalism was mooted as a potential solution to the seemingly intractable conflict between the white and black. In the 1990s, federalism featured heavily as South Africa negotiated a way out of Apartheid and into constitutional democracy. The Constitution ended up being a formally federal one with many substantive unitary features. After decades of these unitary features outshining the federal form of the Constitution, leading to centralisation in the hands of a largely corrupt and inept central governing elite, breathing life into the federal dimensions of the Constitution is once more on the national agenda.
South Africa’s neighbours in the region do not have the same federalist tradition, but for a brief experiment between 1953 and 1963 when the British colonies of Southern Rhodesia (Zimbabwe), Northern Rhodesia (Zambia), and Nyasaland (Malawi) amalgamated into the Federation of Rhodesia and Nyasaland, also called the Central African Federation. What stands out about this experiment is that it was federal in name only, but rather than being unitary in practice, was in fact confederal, a seemingly rare phenomenon. Malawi and Zambia, on the eve of their independence, insisted upon the right to secede, which the British agreed to but which was resented by the white political class of Southern Rhodesia. Regrettably this experiment did not seem to result in any kind of federalist thought or activism, and is largely forgotten today. Namibia was administered as a part of South Africa between 1915 and 1990, and as such to whatever extent it can said to have a federalist tradition shares it with South Africa. Before 1966, Bechuanaland (Botswana), Swaziland, and Basutoland (Lesotho), formed the High Commission Territories, being under the direct control of the British High Commissioner in Cape Town.
South Africa, Zimbabwe, Zambia, Malawi, Botswana, Swaziland, and Lesotho were all states directly or indirectly governed or significantly influenced by the British Empire and government. Of the remaining Southern African states, Mozambique and Angola were Portuguese possessions before 1975, and Madagascar was a French colony prior to 1960. There are significant historical, legal, and political ties between the states that were part of the British Empire, but economic ties exist between all of Southern Africa.
Federalising Southern Africa?
Key role of federalism in Africa
A substantive federalism holds a greater deal of benefit and promise for Africa than it might do elsewhere, particularly in the West. As David Horowitz argued in the 1990s during South Africa’s political transition, “Federalism generally remains only the wisdom of hindsight in Africa.” What he meant by this is that many of Africa’s contemporary problems would have been absent or of a significantly lesser intensity had African states adopted and embraced federalism eagerly when they were granted independence from colonial powers.
In the West, the notion that the government must recognise the equal legal subjectivity of all those who live under its jurisdiction has become entrenched. The United States federal government, for example, as a portion of the political elite, has no particular “score to settle” with any section of the American population. This does not mean that there is no kind of political repression and victimisation in the West, but rather that this would be exceptional. It is rare to read or hear about the countries traditionally regarded as Western today engaging in repression of political minorities. Instead, it is far more likely that the discourse in these states is concerned with protecting and empowering minorities.
In Africa, even often in quite developed states like South Africa and Egypt, to regard government as a score-settler is not unusual. This is partly due to historically entrenched political oppression (and hence mobilisation) along primarily ethnic and but also religious lines.
If a particular ethnicity is victimised, when the political party with its origins in that ethnic community takes power sometime in the future, it would tend to have a score to settle with the former victimisers of its people. The notion that that party must cease being a representative of only a portion of the population and act as the government of all might be acknowledged formally but tends to take a background role.
In Africa’s prized constitutional democracy, South Africa, this was for instance manifested in 2018 when President Cyril Ramaphosa, perceived to be a cosmopolitan reformer, motivated amending the Constitution to allow government to confiscate private property without needing to pay compensation to owners. Ramaphosa said doing this was about addressing the “original colonial sin” of land dispossession that continues to undermine the well-being of “our people.” Here Ramaphosa, 57 years after South Africa became a fully independent state (87 years after substantial independence), and 24 years after it became a democratic state, is still holding the conduct of colonial settlers from centuries ago over the heads of South Africa’s permanent, politically impotent white minority, and appearing to be concerned exclusively with the ethnic constituency of the African National Congress rather than the population as a whole. In Ramaphosa’s mind, at least implicitly, the South African government has an outstanding score to settle.
While South Africa is formally a federation, it is governed unitarily, which allows the central government the leeway to ride roughshod over the interests and freedom of legal subjects with minimal checking from other spheres of government.
Zimbabwe is perhaps a more potent example of score-settling as a feature of governance. After a protracted low-intensity conflict between the white-dominated (but not monopolised) regime of Rhodesia in the 1960s through 1970s, Robert Mugabe and a formation of Marxists-Leninists were elected to government in 1980, starting an administration that persists to this day. This administration began almost immediately with an ethnic cleansing, as Mugabe sought to eliminate the supporters of a rival liberation movement in the south of the country. With the direct support of North Korea and the implicit support of the white officers’ corps in his military, Mugabe slaughtered around 20,000 ethnic Ndebele. The rest of the 1980s and 1990s were more or less stable, as the British government – the former colonial authority – had agreed to assist Mugabe’s regime financially. When the British refused to continue Mugabe’s increasingly anti-property policies in the mid-1990s, the government jettisoned caution. “Veterans” from the civil conflict against the Rhodesian regime insisted that white-owned land was theirs by right, and with the support of the government seized thousands of private farms owned by white Zimbabweans. With his anti-constitutional policy leading to the failure of a financial sector dependent upon stability in agriculture, and inviting widespread international condemnation, Mugabe’s government became truly unhinged, printing the Zimbabwean dollar into valuelessness and causing total economic collapse.
Other African states have also adopted federalism, Nigeria being perhaps the most prominent. Nonetheless, these federations – like South Africa’s – are often superficial. The “score-settling” nature of governance has a toxic effect on standard federalism’s effective functioning, because if the central government – which usually controls the constitutional court – is captured by a political formation with a score to settle, it will inevitably chip away at the features of federalism until only the descriptor “federation” is left. This implies that the ordinary model of federalism requires some significant level of modification to be more appropriately applicable to the African experience.
Caution and Mindfulness
Federalism is evidently part of the solution to score-settling politics in Southern Africa. What are the chances that Zimbabwe could successfully have committed the Ndebele genocide if the Ndebele of southern Zimbabwe controlled one or several provincial governments complete with their own police forces and militias? Could the “veterans” have freely seized farms if decentralised municipal police forces came to the defence of their farming communities? It is conceivable that either of these could have led to civil war, but at least the victimised minorities would have had a fighting chance. It is by no means guaranteed that downward federalism would have avoided ethnic conflict, but it would likely have diminished its likelihood or at least created a more equal playing field.
However, not only downward federalism could have played a positive role in Zimbabwe, South Africa, or any other African state where political centralisation risks the freedom of action or property of legal subjects. Upward federalism – that is, an existing sovereign state federalises with another, creating a common federal authority between them – could also have had a diffusing effect on political power. Indeed, had the Central African Federation still existed, what are the chances that Mugabe, as the governor of the Zimbabwean state, would have gotten away with an ethnic cleansing? Again, it is not inconceivable, but the likelihood appears to be diminished. More general (that is, impersonal) government tends to mean the government does not have a specific bone of contention with a specific minority, because it is necessarily composed of different interest groups.
This all being the case, the federalisation of polities should never be undertaken for its own sake. If the conclusion is reached that federalisation would be worse for the liberty of individuals and communities in Southern Africa than the states remaining separate sovereignties, the enterprise should not be undertaken. This is, however, not currently the way these questions are being approached. If federalisation is to be avoided, it will likely be due to the respective governments’ vested interests in sovereignty, and if it is to be adopted, it will likely be due to the respective governments’ perceived benefits from unification.
These are the practical obstacles to federalisation, not only in Southern Africa but throughout the continent. African states jealously guard their sovereignty, not based on any deep commitment to the “national freedom” to “self-determine,” but because every sovereignty allows for the existence of a political elite. A reduction in a raw number of sovereignties means the number of political elites will likely also be reduced or decline in prestige.
Another potential obstacle to federalisation is more of a global phenomenon, in the form of a preoccupation with grievance. Grievance politics has always existed, with the United States’ Declaration of Independence very explicitly listing its grievances with Britain. But today grievance has reached the level of constitutional thought. Whereas the Declaration of Independence did contain grievances, the United States Constitution does not. It was a plain – but not uncontentious – recognition of constitutional best practice and the pre-existing natural rights of, at the time at least, enfranchised white men.
The South African Constitution of 1996 became a global standard-setter in grievance constitutionalism. Other than the Preamble, the Constitution contains multiple provisions making reference to the injustices of Apartheid that must be borne foremost in mind by the new political class in its governance of the country. This has rendered the Constitution a permanently “post-Apartheid” one, whether it is the year 2000, 2023, or 2100. It has entrenched grievance.
In 2022 Chile narrowly avoided adopting a new constitution that would have amounted to a snapshot of the progressive agenda in the contemporary Western culture war. It would have contained such things as a right to “respect for neurodiversity,” to a “dignified death,” to “create social media,” and to a “digital space free of violence.” While the proposals were defeated by a vote of 62% to 38%, they did represent a wider problem with constitutional thinking.
It might well be that if Southern Africa were to federalise, the constitution drafted for such a union would in various respects be problematic. Using the South African Constitution as a template, it might well contain an expanded laundry list of “rights” that are not rights, properly conceived, and that the limited taxpaying base will likely not be able to sustain. Furthermore, such a draft constitution, like the present “federal” South African Constitution, might seek to concentrate an unnecessary amount of power at the centre.
If any of this is on the agenda, it is better to avoid federalisation entirely.
Bare Minimums
There could be an acceptable arrangement for federalising Southern Africa, however. I explore some of these here.
Disunified federal authority
A “federal government” in Southern Africa need not, and likely ought not, be a single unified structure like most federal governments are today. Federalism tends to emphasise vertical decentralisation but horizontal decentralisation deserves much attention as well, beyond the trias politica.
The governing principle must be polycentricity, which is taken to refer to overlapping, competing, and adjoining sovereignties, on both the vertical and horizontal planes. The federal constitution should endeavour to recognise multiple centres of power – and allow the creation of more – that would include the executive, legislature, and judiciary, but also extend to civil society formations like organised religion, professions, and the academy.
There should, in other words, not simply be the same old three branches model of government, but additional branches should be added. A revenue and taxation branch is considered separately below. In the Southern African context, a branch dealing with oversight (and the concomitant combatting of corruption) and a branch dealing with elections would also make sense. Since these would not merely be agencies of the executive or committees of the legislature, they would, with some luck, not be subject to the same political distortions of their core mandates.
As far as the executive is concerned, given that it would likely remain the most powerful branch of government anyway, significant internal decentralisation must be brought about. The Swiss model of various portfolios, each independently elected, with a rotating chief executive, is a worthy candidate for consideration. More than that, each executive department should not simply be subject to the ultimate authority of their chosen head, but might include an internal council or commission that comprises delegates from subcentral spheres of government. The federal department of defence, for example, could include a defence council that is staffed by delegates chosen by each constituent member of the federation, and perhaps even delegates from the provinces of those constituent members. Without the cooperation of this council, the federal defence minister should be powerless.
The legislature, too, probably should not be a unified one. It might be worthwhile to have separate legislatures (not merely separate chambers within a single legislature) that deal with separate domains of life. South Africa briefly flirted with this thought – granted, for pernicious racialist reasons – in the late 1970s, proposing the creation of three separate parliaments, each of which would have elected its own cabinet and prime minister, under a single head of state. These parliaments, cabinets, and prime ministers would have been responsible for the “own affairs” of the racial-national group they ostensibly represented.
There is a kernel of a good idea present here. However, rather than separate legislatures for racial groups, it might be worthwhile having separate legislatures, cabinets, and prime ministers for groups comprised voluntarily around culture, language, profession, religion, etc. While each legislature would have its own domain, in the spirit of checks and balances, it might be thought worthwhile to still be necessary for each legislature to approve every other legislature’s proposed bills before they can become (general) law.
Nationalism, and local and regional autonomy
Nationalism is a perfectly legitimate phenomenon that is unavoidable. People who share cultures and languages, and often ethnicities, will tend to wish to congregate and organise themselves politically. Seeking to stamp out nationalism would be rightly seen as an unjustifiable attack on freedom of association, particularly in the Southern African context.
Not only should local and regional autonomy be preserved in accordance with the federal and subsidiarity principles, but it should be expanded. In recognition of the decentralist spirit of federalisation, a Southern African federation should likely recognise the autonomy of Southern African groups that are presently, under the various sovereign states, denied self-determination. This does not necessarily have to be a recognition of territorial autonomy but could be on a corporate basis.
Most contentiously, perhaps, it seems important for the constituent parts of the federation to be allowed to form and maintain their own police forces and militias. A significant aspect of federalism is that it must serve to protect people from government abuse. This would not be possible unless the various spheres of government are armed and ready to do so, should a different sphere seek to flex its tyrannical muscles.
Recognition of rights
A federal constitution for Southern Africa would need to include some recognition of liberty and rights. Indeed, the core directive of government is the protection and advancement of the individual’s freedom of action and right to property. However, a “bill of rights,” as alluded to above, might be a risk in the contemporary age of constitutional thought.
Two more modest proposals are submitted as alternatives.
The first takes inspiration from the privileges and immunities clause of the United States Constitution. In this respect, the federal constitution should simply provide that the citizens of any constituent part of the federation are entitled to all the same rights and freedoms as citizens of any other constituent part. It would be up to the courts of the federation to make sense of how this plays out in practice, especially in light of conflicting rights regimes across the federation.
The second proposal derives from the idea of “common” or “basic justice.” This is to say that, instead of a bill of rights, the federal constitution obliges the federal government to protect and advance the freedom of the individual and protect their property. While many on the left and right believe in a more substantive conception of justice, most of them agree that this is the bare minimum of what a government should be doing.
The various spheres of government must have the power to nullify the laws of any other sphere applicable within their jurisdictions to protect citizens from tyranny.
Democracy
A more nuanced conception of democracy, over the standard one-man-one-vote model of universal franchise, would be necessary when a heterogenous federation comes into being. This is also necessary to diffuse the “score-setting” approach to governance.
It should, ideally, not matter that a constituent part of the federation has a greater population than any other. Mere numerical strength must be constitutionally discounted from electoral relevance. Various minority vetoes and confirmatory regional referendums must be considered. The election of important federal officers should likely also be done by way of electoral college, to which constituent parts of the federation send equal numbers of delegates.
Taxation
The issue of taxation in the Southern African context cannot go unaddressed.
This is because the South African individual and commercial taxpayer cannot be expected to foot the bill for the new Southern African polity, which would ordinarily be the case. Nor, as has been explained, should federalism operate as a mechanism for centralisation. As such, a creative solution to the pre-existing problem of over-taxation in the region should be devised.
One of them is to consider the creation of an additional (to the trias politica) independent branch of government that deals not only with the collection of revenue but also with the determination of tax rates.
There is no tax base in the region that can pay for a new federal government, alongside the various existing national, provincial, and municipal governments. As such, there should rather be a single tax structure. This should not be confused with the centralisation of taxation under the federal government. Instead, the new independent “fourth branch” of government should have a constitution reliant on all vertical spheres of government. This means each locality, each province, and each state, should have a significant say in the determination of the single tax rate that applies to all taxpayers.
Taxpayers should also be granted a measure of discretion about where their tax money goes. The federal constitution, then, could determine that half of all the taxes a taxpayer is required to pay goes to the destination – federal, state, provincial, or local – chosen by the taxpayer, whereas the remaining half is spread out by the revenue branch. The revenue branch cannot simply be an expression of the will of the majority in the legislature. Significant veto powers must vest at both the national and potentially provincial and municipal spheres.
Conclusion
Whether one is to support the federalisation of Southern Africa is contingent upon several factors. The most important of these factors is that the constitution proposed for the federation must be a proper federal constitution that emphasises subsidiarity and limited government. The role of the federal government – indeed any government – must primarily be to ensure that the individual’s freedom of action and property are recognised and respected by subcentral governments.
Martin van Staden is an author, jurist, and policy commentator active in South Africa. Send him mail.
Isonomia Quarterly 