Back to the bad old days of the OAU? Why there has never been a Pan-African consensus on the legitimacy of coups d’état

The recent wave of West African military coups threatens to undermine African Union (AU) and Economic Community of West African States (ECOWAS) prohibitions on ‘unconstitutional changes of government.’ Well-connected incomers have always been able to shape how these rules are applied. Unelected leaders in Chad and Zimbabwe, for example, have managed to escape harsh sanctions such as those imposed more recently on the Alliance of Sahelian States (AES). And Burkina Faso’s transitional government after the 2014 insurrection did manage to persuade the AU and ECOWAS that popular uprisings could be tacitly distinguished from coups. But urgent security imperatives in the Sahel are now being used to argue, successfully, for an explicit dilution of these norms. In September 2023, the EU announced that its position on Niger was ‘clear and unchanged: the EU does not recognise authorities emanating from the putsch’. More recently an increasing number of international partners have begun to refer to the deposed Bazoum government in the past tense. 

Some lawyers and scholars have argued, however, that accepting this new reality would represent a return to the ‘bad old days’ of the OAU (1963-2002). This was a time, allegedly, when African states unanimously recognised (violent) coups as legitimate means of taking power. The AES, meanwhile, argues that sanctioning of military regimes is a betrayal of the sovereigntist vision of the OAU’s ‘great Pan-African leaders’. Both sides to this argument, therefore, share a certain understanding of the Pan-African past. For the AES, however, the bad old days were in fact the good old days. 

My main aim here is to show why this accepted version is bad history. The OAU did not simply agree to deal with whoever controlled the governmental apparatus. While it is true that democracy was not proposed as necessary for regime recognition until after 1997, when the Armed Forces Revolutionary Council took control of Freetown, there had been never been a stable consensus on de facto control as the appropriate legal test. Particular controversy surrounded two points. Firstly, whether states could make sovereign decisions about recognition, or whether a common Pan-African position was required. Secondly, whether governments that were in some sense foreign puppets could be denied legitimacy. There were at least two ways of approaching that second question. States styling themselves as anti-colonialist typically assessed new regimes with a language of informal imperialism inherited from Hobson and Lenin via Nkrumah. Others stuck more closely to an established international legal argument, dating from the mid-19th century, that legitimacy could be denied to regimes that had been installed through primarily foreign-sponsored coups. 

These controversies pre-date decolonisation. In early 1958 Liberia persuaded incoming elites to commit to abstaining from interference in all other African states, including those (like itself) that had signed defence pacts with non-African powers. By the end of the year, however, the radical Casablanca group – and most notably Sekou Touré’s Guinea and Kwame Nkrumah’s Ghana – had identified even countries with foreign military bases as potentially illegitimate puppets. The situation complicated dramatically in 1963, in the run-up to the first OAU summit. Togolese President Sylvanus Olympio was assassinated during Africa’s first military coup. Both then and now, there were widespread suspicions of either French or American involvement. Nkrumah, however, enraged Touré and the other Casablanca states by immediately recognising the new Grunitsky government. Thanks to Kate Skinner we now know that Nkrumah, who had been threatening to annexe Togo, had persuaded Grunitsky to sign a secret treaty annexe ceding significant sovereignty to Ghana. His actions ended up strengthening the view that recognition should be a collective not individual matter. 

At its second summit conference in 1964 the OAU split again, with the Casablanca group refusing to allow Congolese PM Moises Tshombe to attend because his Katanganese secessionists had employed white mercenaries. When Nkrumah was deposed in 1966 Guinea, Mali, Tanzania, and Kenya withdrew from the conference rather than sit alongside the incoming National Liberation Council, which the first three of those countries considered to be a Western puppet. In 1971 the deadlock over whether to recognise Idi Amin as Ugandan dictator was so bad that none of the 32 agenda items for that conference could even be discussed. (Ironically, in retrospect, Amin’s opponents alleged British support for his coup.) In 1974-5 it looked as if the MPLA might not be recognised as the legitimate government of Angola. Zambia’s Kenneth Kaunda, in particular, was keen to pursue détente with apartheid South Africa, and saw the brokering of a settlement between UNITA, the MPLA and the FNLA as a means of containing communist influence in the region. The most popular legal justification for this position was that the presence of Cuban troops in Angola made the MPLA a product of foreign subversion. It was argument that nearly succeeded. In January 1976, with South Africa now invading Angola, the OAU voted 22 in favour of a unity government and 22 in favour of recognising the MPLA. When Chairman Idi Amin cast his deciding vote for the socialist camp South Africa withdrew. Senegal wouldn’t recognise the authorities in Luanda until 1982. 

The point here is that there was no thickening consensus after early uncertainty, as is often the case with new international institutions. De facto control never quite functioned as a norm governing regime recognition. Nor was this problem only a product of Cold War paranoia. Samuel Doe’s notorious People’s Redemption Council struggled for recognition in Liberia throughout its ten-year existence. Indeed, when Doe was himself overthrown by Charles Taylor in 1990 (with Libyan, Ivorian, and Burkinabe support) – Nigeria and ECOWAS effectively recognised and propped-up a third rebel faction rather than deal with either the de jure or de facto government. As Kofi Kufour has argued, this refusal to recognise can probably be explained by a fear amongst elderly incumbents that with Doe’s People’s Redemption Council a ‘new class a coup-plotters had burst onto the scene.’ Junior officers, often versed in the language of Afro-Marxism, were now challenging a complacent gerontocracy. These older leaders’ fears fed on partial understandings of the Emperor Haile Seliasse’s assassination by the Derg in Ethiopia and were strongly reinforced by the informal alliance between Jerry Rawlings and Thomas Sankara that emerged after 1983. It may even be, as Frederick Cowell has suggested, that West Africa’s apparent striking enthusiasm for the new constitutional governance norms of the 1990s was also a cynical means of coup-proofing regimes against these threats from the ‘bottom-up’.  

We should be sceptical, therefore, of the claim by West Africa’s new military governments that recognising them would represent a return to a Pan-African respect for sovereignty and a rejection of neo-colonialism. Combining these concepts actually helps opens to the door to endless arguments about the legitimacy of new regimes, such as those that dogged the OAU throughout its existence. Arguments about foreign support could, moreover, be easily turned against the AES themselves (given their preference for Russian over French and Amercian support). There are, finally, good reasons to think that the old question of foreign-sponsored coups is likely to prove particularly persistent in West Africa. The practical constraints on statebuilding remain ferocious, and social depth has proved elusive for all governments, even those more firmly in control of their territory than the AES is today. As Jean-Pierre Olivier de Sardan has recently argued, West African civilian and military administrations have always loudly accused the other of laxism and corruption without ever consistently demonstrating any improved capacity to solve these problems. With domestic support being so difficult to secure, even small amounts of external assistance may continue to facilitate the seizure and even maintenance of power. 

Photo: Kassim Traore (2020)