The African National Congress has something of a problem with regard to leadership transitions. Of four democratic presidents of South Africa, it looks likely that only one – Madiba – will have completed his term as state president. The other two, Presidents Thabo Mbeki and Jacob Zuma, have been asked to step down early by their own parties, while Kgalema Motlanthe served essentially as a caretaker president. Will it be that all ANC presidents will be unlikely to finish their term as state president?

Part of the reason for this is the ANC’s alignment of its national conference, which elects its new president (and choice for state president) roughly 18 months before the national election. We’ll come back to this.

The bigger political question raised by all of this drama is that of managing leadership transition.

What is political power? Political scientists can debate this endlessly, but for our purposes, let us define it as the ability to influence political events. It is the ability to decide what will happen.

Political power – or political capital – waxes and wanes. A new party leader, having just received a mandate from his party and seen off one or more competitors, is at the height of his or her power. Having been chosen to lead their party, to wield power on its behalf, they carry tremendous authority. A major part of a leader’s power flows from his ability, as the (incoming) head of the executive, to decide what government will do, which has of course enormous social and economic implications.

As, or even more, important, as someone who will hire and fire ministers, ambassadors and others, and influence myriad influential public appointments, he has the power to affect the careers and livelihoods of the political class over which he presides. This is one of the biggest sources of power for political leaders generally, but it is even more important in African politics when many political elites have far fewer options than in mature democracies to achieve a comparable livelihood or status outside of political office.

What happens when this power wanes? It is important to answer this question because South Africa, as an aspiring developmental state, needs political power to be exercised effectively. We do not have the luxury of having ineffectual, lame duck presidents in office for months and years.

Arguably, President Zuma’s power began to wane significantly when the ANC performed poorly in the 2016 local government elections; 2017 saw further political drift as it became clear that a significant section of the party wanted to be rid of him, but a stalemate position emerged that removing him before the national conference in December would be more trouble than it was worth.

Now it is widely accepted that President Zuma finishing his term as state president is unthinkable if the ANC wants to claw back some of the support and legitimacy lost during Zuma’s time at the helm of the party. The only question is when.

President Zuma’s second term has thus exposed at least two under-explored deficiencies in our politics. First, the misalignment of party and state leadership terms both hastens the lame duck period of a state president’s term and creates the much talked-about “two centres of power” problem. Second, we don’t have effective political norms for removing a disastrous leader before they complete their term.

On the first problem, we saw a significant loss of political and economic momentum in 2017 as political actors concerned themselves with campaigning, choosing sides and manoeuvring ahead of the ANC’s elective conference. Investors and business adopted a wait-and-see approach to investment, uncertain of who would emerge to run the country, and what their agenda would be on key economic policy questions. This kind of uncertainty is undesirable only two-and-a-half years into the five-year mandate of an administration.

Of course, post-Nasrec, we have also had the re-emergence of the “two centres of power” problem. Even in the best case – where an ANC president is taking over from a predecessor he does not have major political differences with – it is still undesirable to have to manage two centres of power for 18 months. As we’ve established, political power is the ability to decide what will happen, on policy, appointments, and the hundreds of consequential decisions which are made in the course of running a country. Who decides? If the state president doesn’t have the last word on government decision-making because he has to consider the preferences of his party president, why stay in the chair? Our national experience in the democratic era has been that this hasn’t been managed well, and creates unnecessary complication.

It is worth looking at how more mature democracies manage this situation. In the United States, a sitting president’s successor is elected only two months before he leaves office. The “two centres of power” scenario is thus mercifully brief for all concerned. In the UK, a parliamentary democracy with fundamental similarities to our own, a more decisive convention is applied where the party leader immediately takes over as prime minister. The same holds true for many other comparable parliamentary democracies, such as Germany and Israel.

Political conventions tend to persist due to the sheer inertia of the status quo, as well as the extent to which powerful stakeholders are invested in it. On the latter, it is unclear who would be invested in ensuring that for as long as the ANC is the ruling party, a two centres of power dynamic will occur for up to 18 months, every five or 10 years.

The obvious solution would be to push back the ANC’s national conference by one year. One counterargument which has been raised to me is that the ANC needs the time post-conference to gear up for the next election. I would counter that the party should maintain dedicated internal capacity for election campaigning, which is merely augmented by the incoming leadership’s vision. Some election preparation could therefore take place concurrently during the run-up to the conference. An incoming leadership team could have up to six months to campaign between a pushed-back ANC conference and the next national election, which seems manageable to me.

Of the many intractable problems in our politics, the two centres of power one seems to me to be self-created, unnecessary and solvable.

The second problem is far more interesting to me. As we have said, it is worrying that the ANC is developing a trend of removing its leaders before their terms as state president have been completed. Removing President Mbeki now looks to have been a mistake, both because the pretext – Mbeki was accused of using state institutions to unfairly target Zuma for prosecution – has since been discredited and is now painfully ironic given the subsequently legally and ethically troubled Zuma presidency. This is not to even mention the reduced majority which resulted, due mainly to the splinter party (COPE) created by members disgruntled by, and disillusioned with, the aftermath of Polokwane.

Paradoxically, having removed one state president when it probably shouldn’t have in 2008, the ANC then failed to remove a state president when it probably should have in 2016 when, variously: former Deputy Minister of Finance, Mcebisi Jonas, made explosive allegations that the President’s close friends – the Gupta family – attempted to corrupt him; the Constitutional Court rebuked the president on his handling of the years-long Nkandla saga, and the ANC performed dismally in the local government elections in a clear rebuke by voters of Zuma’s leadership.

Why didn’t the ANC remove Zuma from power then, despite the obvious damage he was doing to the party, and therefore the political prospects of the ANC leaders who could have acted but did not?

Lest we forget, President Zuma presided over four elections in which the ANC lost electoral support each and every time, as well as control of three metros including the economic hub of Africa in Johannesburg.

In pondering the question of why his party did not act sooner, my thoughts keep coming back to Britain in which parties regularly remove leaders when they are deemed to be past their sell-by dates.

British political commentator Steve Richards has identified three “iron laws”, or conditions which must be fulfilled for a British party to remove its leader:

“There has to be at least one popular alternative candidate. The risk of the coup must be nowhere near as high as sticking with the leader. The removal of the leader must not threaten to generate a civil war.”

The case to remove President Zuma in 2016 fulfilled all of these conditions. The same presidential candidates in December 2017 were available the year before. The ANC was bleeding votes and legitimacy under Zuma, making keeping him obviously worse for the party than removing him. While a split from an exodus of Zuma loyalists was a possibility, it’s hard to argue the party would have been better off being saddled with his political baggage than exorcising itself of it.

By not acting then, the ANC lost the opportunity to demonstrate to voters it could self-correct before being forced to.

Thus the ANC’s mistake was not removing President Zuma in 2016. There were several reasons why this did not happen, in defiance of the iron laws.

The NEC, which had the power to remove Zuma, was cowed by him. The majority of the NEC either relied on him for employment (as ministers), patronage or protection. Or they feared he had information which could hurt them.

Even for those who were not cowed, there was no consensus on whether removing him was the best course for the party. Some feared handing the opposition a victory in their crusade against Zuma. Some worried about the precedent and embarrassment of removing two successive state presidents prematurely. Some worried about whether the party could successfully convene a special conference amid fierce factional battles. Some clung to notions of comradely loyalty to the ANC president, and collective leadership which transcends personalities: noble principles which are baked into the party’s DNA, but unfortunately had the practical effect of protecting an obviously errant leader from accountability.

Perhaps the key difference between a British party applying the iron laws and our own situation is that most ANC leaders, irrespective of faction, do not in their heart of hearts believe that the ANC can lose a national election in the foreseeable future, even despite the erosion of the party’s share of the vote. This affected their calculus of the risk of sticking with an unpopular leader.

Finally having boosted the opposition with COPE in 2009 and EFF in 2014, there was heightened risk aversion to action which could have caused another split.

Some lessons should be learnt from all of this.

The leadership collective of a party needs to be able to hold party leaders accountable, not be hostages to them. Achieving this will require a change of mindset and political culture, while also acknowledging and counteracting the power leaders are able to wield over their collectives.

The Conservative Party in the UK responds to this asymmetry of power in a particularly interesting way, by allowing party leaders to trigger a vote of no confidence anonymously. This forces the party leader to be consultative and maintain the consent of the collective rather than preside imperiously and unaccountably over them. Labour’s removal process is more difficult, but still more practical than the ANC’s, which seems to require a consensus in the NEC, an unreasonably high threshold for dissent.

As our politics matures, we should consider our own “iron laws” and refine our mechanisms for removing leaders who are disastrous, delinquent or merely past their sell-by date. British ruthlessness against failing leaders is based on the calculus that if parties don’t act, voters certainly will. DM